Update the Main Branch from certainty

.gitignore

@@ -1,88 +1,29 @@
-# =========================
+<<<<<<< HEAD
-# Python
+# Ignore all contents of these directories
-# =========================
+!**/*.py
-__pycache__/
+/Data/
-*.py[cod]
+/attach/
-*$py.class
+/results/
-.ipynb_checkpoints/
+/enarcelona/
-
+=======
-# =========================
+# 1. Broad Ignores
-# Virtual environments
+/Data/*
-# =========================
+/attach/*
-env/
+/results/*
-env*/
+/enarcelona/*
-venv/
+>>>>>>> Certainty
-.venv/
-enarcelona/
-
-# =========================
-# Secrets
-# =========================
 .env
-*.env
+__pycache__/
-
+*.pyc
-# =========================
-# Patient data / sensitive data
-# =========================
-Data/
-data/raw/
-data/processed/
-data/ground_truth/
-reference/
-
-# =========================
-# Generated results and logs
-# =========================
-results/
-results_edss_benchmark/
-*.log
-
-# =========================
-# Large/generated file types
-# =========================
 *.csv
-*.tsv
+=======
-*.json
+/reference/
-*.jsonl
-*.xlsx
-*.xls
-*.png
-*.PNG
-*.jpg
-*.jpeg
 *.svg
-*.pdf
+>>>>>>> Stashed changes
+# 2. Ignore virtual environments COMPLETELY
+# This must come BEFORE the unignore rule
+env*/
 
-# =========================
+# 3. The "Unignore" rule (Whitelisting)
-# Temporary / backup files
+# We only unignore .py files that aren't already blocked by the rules above
-# =========================
+!**/*.py
-*.tmp
-*.bak
-*.orig
-.DS_Store
-
-# =========================
-# Keep important code/config/docs
-# =========================
-!README.md
-!requirements.txt
-!*.py
-!*.md
-!*.yml
-!*.yaml
-!*.toml
-
-# Keep prompt templates / schemas if safe to publish
-!prompts/
-!prompts/**
-!attach/
-!attach/*.gbnf
-!attach/just_edss_text.txt
-!attach/Komplett.txt
-
-# Keep example/synthetic data only
-!data/
-!data/example/
-!data/example/**
-!Data/example/
-!Data/example/**

.gitignore.orig

@@ -0,0 +1,29 @@
+<<<<<<< HEAD
+# Ignore all contents of these directories
+!**/*.py
+/Data/
+/attach/
+/results/
+/enarcelona/
+=======
+# 1. Broad Ignores
+/Data/*
+/attach/*
+/results/*
+/enarcelona/*
+>>>>>>> Certainty
+.env
+__pycache__/
+*.pyc
+*.csv
+=======
+/reference/
+*.svg
+>>>>>>> Stashed changes
+# 2. Ignore virtual environments COMPLETELY
+# This must come BEFORE the unignore rule
+env*/
+
+# 3. The "Unignore" rule (Whitelisting)
+# We only unignore .py files that aren't already blocked by the rules above
+!**/*.py

PROJECT_STRUCTURE.md

@@ -1,31 +0,0 @@
-# Project Structure
-
-This project was reorganized into:
-
-- `data/`
-  - `raw/`: original raw data, if retained locally
-  - `processed/`: cleaned or derived input data
-  - `ground_truth/`: manually annotated reference data
-  - `external/`: externally provided data
-
-- `prompts/`
-  - EDSS instructions and prompt/schema assets
-
-- `scripts/`
-  - runnable analysis and plotting scripts
-
-- `results/`
-  - `benchmark_runs/`: full model benchmark runs
-  - `final_results/`: final selected model outputs
-  - `figures/`: generated figures
-  - `tables/`: generated tables
-  - `logs/`: terminal logs
-
-- `manuscript/`
-  - final figures and tables for paper/thesis writing
-
-- `archive/`
-  - old scripts, old results, temporary files, and unclear legacy files
-
-Important:
-The reorganization was performed after creating a full timestamped backup.

app.py

@@ -216,3 +216,6 @@ if __name__ == "__main__":
 
 
 
+# %% name
+eXXXXXXXX
+##

certainty.py

@@ -0,0 +1,600 @@
+
+# %% API call1
+#import time
+#import json
+#import os
+#from datetime import datetime
+#import pandas as pd
+#from openai import OpenAI
+#from dotenv import load_dotenv
+#
+## Load environment variables
+#load_dotenv()
+#
+## === CONFIGURATION ===
+#OPENAI_API_KEY = os.getenv("OPENAI_API_KEY")
+#OPENAI_BASE_URL = os.getenv("OPENAI_BASE_URL")
+#MODEL_NAME = "GPT-OSS-120B"
+#HEALTH_URL = f"{OPENAI_BASE_URL}/health"  # Placeholder - actual health check would need to be implemented
+#CHAT_URL = f"{OPENAI_BASE_URL}/chat/completions"
+#
+## File paths
+#INPUT_CSV = "/home/shahin/Lab/Doktorarbeit/Barcelona/Data/MS_Briefe_400_with_unique_id_SHA3_explore_cleaned_unique.csv"
+#EDSS_INSTRUCTIONS_PATH = "/home/shahin/Lab/Doktorarbeit/Barcelona/attach/Komplett.txt"
+##GRAMMAR_FILE = "/home/shahin/Lab/Doktorarbeit/Barcelona/attach/just_edss_schema.gbnf"
+#
+## Initialize OpenAI client
+#client = OpenAI(
+#    api_key=OPENAI_API_KEY,
+#    base_url=OPENAI_BASE_URL
+#)
+#
+## Read EDSS instructions from file
+#with open(EDSS_INSTRUCTIONS_PATH, 'r') as f:
+#    EDSS_INSTRUCTIONS = f.read().strip()
+## === RUN INFERENCE 2 ===
+#def run_inference(patient_text):
+#    prompt = f'''
+#    Du bist ein medizinischer Assistent, der spezialisiert darauf ist, EDSS-Scores (Expanded Disability Status Scale) aus klinischen Berichten zu extrahieren.
+#### Regeln für die Ausgabe:
+#1. **Reason**: Erstelle eine prägnante Zusammenfassung (max. 400 Zeichen) der Befunde auf **DEUTSCH**, die zur Einstufung führen.
+#2. **klassifizierbar**:
+#   - Setze dies auf **true**, wenn ein EDSS-Wert identifiziert, berechnet oder basierend auf den klinischen Hinweisen plausibel geschätzt werden kann.
+#   - Setze dies auf **false**, NUR wenn die Daten absolut unzureichend oder so widersprüchlich sind, dass keinerlei Einstufung möglich ist.
+#3. **EDSS**:
+#   - Dieses Feld ist **VERPFLICHTEND**, wenn "klassifizierbar" auf true steht.
+#   - Es muss eine Zahl zwischen 0.0 und 10.0 sein.
+#   - Versuche stets, den EDSS-Wert so präzise wie möglich zu bestimmen, auch wenn die Datenlage dünn ist (nutze verfügbare Informationen zu Gehstrecke und Funktionssystemen).
+#   - Dieses Feld **DARF NICHT ERSCHEINEN**, wenn "klassifizierbar" auf false steht.
+#
+#### Einschränkungen:
+#- Erfinde keine Fakten, aber nutze klinische Herleitungen aus dem Bericht, um den EDSS zu bestimmen.
+#- Priorisiere die Vergabe eines EDSS-Wertes gegenüber der Markierung als nicht klassifizierbar.
+#- Halte dich strikt an die JSON-Struktur.
+#
+#EDSS-Bewertungsrichtlinien:
+#{EDSS_INSTRUCTIONS}
+#
+#Patientenbericht:
+#{patient_text}
+#'''
+#    start_time = time.time()
+#
+#    try:
+#        # Make API call using OpenAI client
+#        response = client.chat.completions.create(
+#            messages=[
+#                {
+#                    "role": "system",
+#                    "content": "You extract EDSS scores. You prioritize providing a score even if data is partial, by using clinical inference."
+#                },
+#                {
+#                    "role": "user",
+#                    "content": prompt
+#                }
+#            ],
+#            model=MODEL_NAME,
+#            max_tokens=2048,
+#            temperature=0.0,
+#            response_format={"type": "json_object"}
+#        )
+#
+#        # Extract content from response
+#        content = response.choices[0].message.content
+#
+#        # Parse the JSON response
+#        parsed = json.loads(content)
+#
+#        inference_time = time.time() - start_time
+#
+#        return {
+#            "success": True,
+#            "result": parsed,
+#            "inference_time_sec": inference_time
+#        }
+#
+#    except Exception as e:
+#        print(f"Inference error: {e}")
+#        return {
+#            "success": False,
+#            "error": str(e),
+#            "inference_time_sec": -1
+#        }
+## === BUILD PATIENT TEXT ===
+#def build_patient_text(row):
+#    return (
+#        str(row["T_Zusammenfassung"]) + "\n" +
+#        str(row["Diagnosen"]) + "\n" +
+#        str(row["T_KlinBef"]) + "\n" +
+#        str(row["T_Befunde"]) + "\n"
+#    )
+#
+#if __name__ == "__main__":
+#    # Read CSV file ONLY inside main block
+#    df = pd.read_csv(INPUT_CSV, sep=';')
+#    results = []
+#
+#    # Process each row
+#    for idx, row in df.iterrows():
+#        print(f"Processing row {idx + 1}/{len(df)}")
+#        try:
+#            patient_text = build_patient_text(row)
+#            result = run_inference(patient_text)
+#
+#            # Add unique_id and MedDatum to result for tracking
+#            result["unique_id"] = row.get("unique_id", f"row_{idx}")
+#            result["MedDatum"] = row.get("MedDatum", None)
+#
+#            results.append(result)
+#            print(json.dumps(result, indent=2))
+#        except Exception as e:
+#            print(f"Error processing row {idx}: {e}")
+#            results.append({
+#                "success": False,
+#                "error": str(e),
+#                "unique_id": row.get("unique_id", f"row_{idx}"),
+#                "MedDatum": row.get("MedDatum", None)
+#            })
+#
+#    # Save results to a JSON file
+#    output_json = INPUT_CSV.replace(".csv", "_results_Nisch.json")
+#    with open(output_json, 'w') as f:
+#        json.dump(results, f, indent=2)
+#    print(f"Results saved to {output_json}")
+##
+
+
+
+# %% API call1 - Enhanced with certainty scoring
+#import time
+#import json
+#import os
+#from datetime import datetime
+#import pandas as pd
+#from openai import OpenAI
+#from dotenv import load_dotenv
+#
+## Load environment variables
+#load_dotenv()
+#
+## === CONFIGURATION ===
+#OPENAI_API_KEY = os.getenv("OPENAI_API_KEY")
+#OPENAI_BASE_URL = os.getenv("OPENAI_BASE_URL")
+#MODEL_NAME = "GPT-OSS-120B"
+#
+## File paths
+#INPUT_CSV = "/home/shahin/Lab/Doktorarbeit/Barcelona/Data/Test.csv"
+#EDSS_INSTRUCTIONS_PATH = "/home/shahin/Lab/Doktorarbeit/Barcelona/attach/Komplett.txt"
+#
+## Initialize OpenAI client
+#client = OpenAI(
+#    api_key=OPENAI_API_KEY,
+#    base_url=OPENAI_BASE_URL
+#)
+#
+## Read EDSS instructions from file
+#with open(EDSS_INSTRUCTIONS_PATH, 'r') as f:
+#    EDSS_INSTRUCTIONS = f.read().strip()
+#
+## === PROMPT WITH CERTAINTY REQUEST ===
+#def build_prompt(patient_text):
+#    return f'''Du bist ein medizinischer Assistent, der spezialisiert darauf ist, EDSS-Scores (Expanded Disability Status Scale), alle Unterkategorien und die Bewertungssicherheit aus klinischen Berichten zu extrahieren.
+#
+#### Deine Aufgabe:
+#1. Analysiere den Patientenbericht und extrahiere:
+#   - Den Gesamt-EDSS-Score (0.0–10.0)
+#   - Alle 8 EDSS-Unterkategorien (mit jeweils eigener Maximalpunktzahl)
+#2. Schätze für jede Entscheidung die Sicherheit als Ganzzahl von 0–100 % ein.
+#
+#### Struktur der JSON-Ausgabe (VERPFLICHTEND):
+#Gib NUR gültiges JSON zurück — kein Markdown, kein Text davor/dahinter.
+#
+#{{
+#  "reason": "Kernaussage zur EDSS-Begründung (max. 400 Zeichen, auf Deutsch).",
+#  "klassifizierbar": true/false,
+#  "EDSS": null ODER Zahl zwischen 0.0 und 10.0 (nur wenn klassifizierbar=true)",
+#  "certainty_percent": 0 ODER Zahl zwischen 0 und 100 (Ganzzahl)",
+#  "subcategories": {{
+#    "VISUAL_OPTIC_FUNCTIONS": null ODER Zahl zwischen 0.0 und 6.0,
+#    "BRAINSTEM_FUNCTIONS": null ODER Zahl zwischen 0.0 und 6.0,
+#    "PYRAMIDAL_FUNCTIONS": null ODER Zahl zwischen 0.0 und 6.0,
+#    "CEREBELLAR_FUNCTIONS": null ODER Zahl zwischen 0.0 und 6.0,
+#    "SENSORY_FUNCTIONS": null ODER Zahl zwischen 0.0 und 6.0,
+#    "BOWEL_AND_BLADDER_FUNCTIONS": null ODER Zahl zwischen 0.0 und 6.0,
+#    "CEREBRAL_FUNCTIONS": null ODER Zahl zwischen 0.0 und 6.0,
+#    "AMBULATION": null ODER Zahl zwischen 0.0 und 10.0
+#  }}
+#}}
+#
+#### Regeln:
+#- **reason**: Kurze, prägnante Begründung (auf Deutsch, max. 400 Zeichen), warum du den EDSS-Wert und die Unterkategorien so bewertest.
+#- **klassifizierbar**:
+#  - `true`, wenn EDSS und mindestens die wichtigsten Unterkategorien *eindeutig ableitbar* oder *plausibel inferierbar* sind.
+#  - `false`, **nur**, wenn keine relevanten Daten vorliegen, oder diese so widersprüchlich/inkonsistent sind, dass keine vernünftige Einschätzung möglich ist.
+#- **EDSS**:
+#  - **VERPFLICHTEND**, wenn `klassifizierbar=true`.
+#  - Zahl zwischen 0.0 und 10.0 (z.B. 3.0, 5.5). Darf **nicht** erscheinen, wenn `klassifizierbar=false`.
+#- **certainty_percent**:
+#  - **Immer present** — Ganzzahl (0–100), basierend auf:
+#    - Klarheit und Vollständigkeit der Berichtsangaben,
+#    - Stichhaltigkeit der Schlussfolgerung (inkl. Inferenz),
+#    - Konsistenz zwischen den Unterkategorien.
+#- **subcategories**:
+#  - **Immer present** — **alle 8 Unterkategorien** müssen enthalten sein.
+#  - Jeder Wert ist entweder:
+#    - `null` (wenn keine ausreichende Information vorliegt), **oder**
+#    - eine Zahl ≤ jeweiliger Obergrenze (z.B. Ambulation ≤ 10.0).
+#  - Wenn die Unterkategorie plausibel inferiert werden kann (auch indirekt), gib einen sinnvollen Wert ab.
+#  - Beispiel: Wenn „Gang mit Krückstock auf ebenem Boden bis 200 m“ steht, setze `AMBULATION: 5.5`.
+#
+#### EDSS-Bewertungsrichtlinien:
+#{EDSS_INSTRUCTIONS}
+#
+#Patientenbericht:
+#{patient_text}
+#'''
+#
+## === INFERENCE FUNCTION ===
+#def run_inference(patient_text):
+#    prompt = build_prompt(patient_text)
+#
+#    start_time = time.time()
+#
+#    try:
+#        response = client.chat.completions.create(
+#            messages=[
+#                {"role": "system", "content": "Du gibst EXKLUSIV gültiges JSON zurück — keine weiteren Erklärungen."}
+#            ] + [
+#                {"role": "user", "content": prompt}
+#            ],
+#            model=MODEL_NAME,
+#            max_tokens=2048,
+#            temperature=0.1,  # Slightly higher for more natural certainty estimation (still low for reliability)
+#            response_format={"type": "json_object"}
+#        )
+#
+#        content = response.choices[0].message.content
+#
+#        # Parse and validate JSON
+#        try:
+#            parsed = json.loads(content)
+#        except json.JSONDecodeError as e:
+#            print(f"⚠️  JSON parsing failed: {e}")
+#            print("Raw response:", content[:500])
+#            raise ValueError("Model did not return valid JSON")
+#
+#        # Enforce required keys
+#        if "certainty_percent" not in parsed:
+#            print("⚠️  Missing 'certainty_percent' in output! Force-adding fallback.")
+#            parsed["certainty_percent"] = 0  # fallback
+#        elif not isinstance(parsed["certainty_percent"], (int, float)):
+#            parsed["certainty_percent"] = int(parsed["certainty_percent"])
+#
+#        # Clamp certainty to [0, 100]
+#        pct = parsed["certainty_percent"]
+#        parsed["certainty_percent"] =max(0, min(100, int(pct)))
+#
+#        # Enforce EDSS rules: if not classifiable → remove EDSS
+#        if not parsed.get("klassifizierbar", False):
+#            if "EDSS" in parsed:
+#                del parsed["EDSS"]  # per spec, must not appear if not classifiable
+#        else:
+#            if "EDSS" not in parsed:
+#                print("⚠️  'klassifizierbar' is true but EDSS missing — adding fallback.")
+#                parsed["EDSS"] = 7.0  # last-resort fallback
+#
+#        inference_time = time.time() - start_time
+#
+#        return {
+#            "success": True,
+#            "result": parsed,
+#            "inference_time_sec": inference_time
+#        }
+#
+#    except Exception as e:
+#        print(f"❌ Inference error: {e}")
+#        return {
+#            "success": False,
+#            "error": str(e),
+#            "inference_time_sec": -1,
+#            "result": None  # no structured output
+#        }
+#
+## === BUILD PATIENT TEXT ===
+#def build_patient_text(row):
+#    return (
+#        str(row.get("T_Zusammenfassung", "")) + "\n" +
+#        str(row.get("Diagnosen", "")) + "\n" +
+#        str(row.get("T_KlinBef", "")) + "\n" +
+#        str(row.get("T_Befunde", ""))
+#    )
+#
+#if __name__ == "__main__":
+#    # Load data
+#    df = pd.read_csv(INPUT_CSV, sep=';')
+#    results = []
+#
+#    # Optional: limit for testing
+#    # df = df.head(3)
+#
+#    print(f"Processing {len(df)} rows...")
+#    for idx, row in df.iterrows():
+#        print(f"\n— Row {idx + 1}/{len(df)} —")
+#        try:
+#            patient_text = build_patient_text(row)
+#            result = run_inference(patient_text)
+#
+#            # Attach metadata
+#            result["unique_id"] = row.get("unique_id", f"row_{idx}")
+#            result["MedDatum"] = row.get("MedDatum", None)
+#
+#            results.append(result)
+#
+#            # Print summary
+#            if result["success"]:
+#                res = result["result"]
+#                edss = res.get("EDSS", "N/A") if res.get("klassifizierbar") else "N/A"
+#                print(f"✅ Result → EDSS={edss}, certainty={res.get('certainty_percent', 'N/A')}%")
+#                print(f"   Reason: {res.get('reason', 'N/A')[:100]}…")
+#            else:
+#                print(f"❌ Failed: {result.get('error', 'Unknown error')[:100]}")
+#
+#        except Exception as e:
+#            print(f"⚠️  Error processing row {idx}: {e}")
+#            results.append({
+#                "success": False,
+#                "error": str(e),
+#                "unique_id": row.get("unique_id", f"row_{idx}"),
+#                "MedDatum": row.get("MedDatum", None),
+#                "result": None
+#            })
+#
+#    # Save results
+#    output_json = INPUT_CSV.replace(".csv", "_results_Nisch_certainty.json")
+#    with open(output_json, 'w', encoding='utf-8') as f:
+#        json.dump(results, f, indent=2, ensure_ascii=False)
+#    print(f"\n✅ Saved results to: {output_json}")
+#
+##
+
+
+# %% API call - Multi-iteration EDSS + certainty extraction
+
+import time
+import json
+import os
+from datetime import datetime
+import pandas as pd
+from openai import OpenAI
+from dotenv import load_dotenv
+
+# Load environment variables
+load_dotenv()
+
+# === CONFIGURATION ===
+OPENAI_API_KEY = os.getenv("OPENAI_API_KEY")
+OPENAI_BASE_URL = os.getenv("OPENAI_BASE_URL")
+MODEL_NAME = "GPT-OSS-120B"
+
+# File paths
+INPUT_CSV = "/home/shahin/Lab/Doktorarbeit/Barcelona/Data/MS_Briefe_400_with_unique_id_SHA3_explore_cleaned_unique.csv"
+EDSS_INSTRUCTIONS_PATH = "/home/shahin/Lab/Doktorarbeit/Barcelona/attach/Komplett.txt"
+
+# Iteration settings
+NUM_ITERATIONS = 20
+STOP_ON_FIRST_ERROR = False  # Set to True for debugging
+
+# Initialize OpenAI client
+client = OpenAI(
+    api_key=OPENAI_API_KEY,
+    base_url=OPENAI_BASE_URL
+)
+
+# Read EDSS instructions from file
+with open(EDSS_INSTRUCTIONS_PATH, 'r') as f:
+    EDSS_INSTRUCTIONS = f.read().strip()
+
+# === PROMPT (unchanged from before) ===
+def build_prompt(patient_text):
+    return f'''Du bist ein medizinischer Assistent, der spezialisiert darauf ist, EDSS-Scores (Expanded Disability Status Scale), alle Unterkategorien und die Bewertungssicherheit aus klinischen Berichten zu extrahieren.
+
+### Deine Aufgabe:
+1. Analysiere den Patientenbericht und extrahiere:
+   - Den Gesamt-EDSS-Score (0.0–10.0)
+   - Alle 8 EDSS-Unterkategorien (mit jeweils eigener Maximalpunktzahl)
+2. Schätze für jede Entscheidung die Sicherheit als Ganzzahl von 0–100 % ein.
+
+### Struktur der JSON-Ausgabe (VERPFLICHTEND):
+Gib NUR gültiges JSON zurück — kein Markdown, kein Text davor/dahinter.
+
+{{
+  "reason": "Kernaussage zur EDSS-Begründung (max. 400 Zeichen, auf Deutsch).",
+  "klassifizierbar": true/false,
+  "EDSS": null ODER Zahl zwischen 0.0 und 10.0 (nur wenn klassifizierbar=true)",
+  "certainty_percent": 0 ODER Zahl zwischen 0 und 100 (Ganzzahl)",
+  "subcategories": {{
+    "VISUAL_OPTIC_FUNCTIONS": null ODER Zahl zwischen 0.0 und 6.0,
+    "BRAINSTEM_FUNCTIONS": null ODER Zahl zwischen 0.0 und 6.0,
+    "PYRAMIDAL_FUNCTIONS": null ODER Zahl zwischen 0.0 und 6.0,
+    "CEREBELLAR_FUNCTIONS": null ODER Zahl zwischen 0.0 und 6.0,
+    "SENSORY_FUNCTIONS": null ODER Zahl zwischen 0.0 und 6.0,
+    "BOWEL_AND_BLADDER_FUNCTIONS": null ODER Zahl zwischen 0.0 und 6.0,
+    "CEREBRAL_FUNCTIONS": null ODER Zahl zwischen 0.0 und 6.0,
+    "AMBULATION": null ODER Zahl zwischen 0.0 und 10.0
+  }}
+}}
+
+### Regeln:
+- **reason**: Kurze, prägnante Begründung (auf Deutsch, max. 400 Zeichen), warum du den EDSS-Wert und die Unterkategorien so bewertest.
+- **klassifizierbar**:
+  - `true`, wenn EDSS und mindestens die wichtigsten Unterkategorien *eindeutig ableitbar* oder *plausibel inferierbar* sind.
+  - `false`, **nur**, wenn keine relevanten Daten vorliegen, oder diese so widersprüchlich/inkonsistent sind, dass keine vernünftige Einschätzung möglich ist.
+- **EDSS**:
+  - **VERPFLICHTEND**, wenn `klassifizierbar=true`.
+  - Zahl zwischen 0.0 und 10.0 (z.B. 3.0, 5.5). Darf **nicht** erscheinen, wenn `klassifizierbar=false`.
+- **certainty_percent**:
+  - **Immer present** — Ganzzahl (0–100), basierend auf:
+    - Klarheit und Vollständigkeit der Berichtsangaben,
+    - Stichhaltigkeit der Schlussfolgerung (inkl. Inferenz),
+    - Konsistenz zwischen den Unterkategorien.
+- **subcategories**:
+  - **Immer present** — **alle 8 Unterkategorien** müssen enthalten sein.
+  - Jeder Wert ist entweder:
+    - `null` (wenn keine ausreichende Information vorliegt), **oder**
+    - eine Zahl ≤ jeweiliger Obergrenze (z.B. Ambulation ≤ 10.0).
+  - Wenn die Unterkategorie plausibel inferiert werden kann (auch indirekt), gib einen sinnvollen Wert ab.
+  - Beispiel: Wenn „Gang mit Krückstock auf ebenem Boden bis 200 m“ steht, setze `AMBULATION: 5.5`.
+
+### EDSS-Bewertungsrichtlinien:
+{EDSS_INSTRUCTIONS}
+
+Patientenbericht:
+{patient_text}
+'''
+
+# === INFERENCE FUNCTION (unchanged) ===
+def run_inference(patient_text):
+    prompt = build_prompt(patient_text)
+
+    start_time = time.time()
+
+    try:
+        response = client.chat.completions.create(
+            messages=[
+                {"role": "system", "content": "Du gibst EXKLUSIV gültiges JSON zurück — keine weiteren Erklärungen."}
+            ] + [
+                {"role": "user", "content": prompt}
+            ],
+            model=MODEL_NAME,
+            max_tokens=2048,
+            temperature=0.1,
+            response_format={"type": "json_object"}
+        )
+
+        content = response.choices[0].message.content
+
+        # Parse and validate JSON
+        try:
+            parsed = json.loads(content)
+        except json.JSONDecodeError as e:
+            print(f"⚠️   JSON parsing failed: {e}")
+            print("Raw response:", content[:500])
+            raise ValueError("Model did not return valid JSON")
+
+        # Enforce required keys
+        if "certainty_percent" not in parsed:
+            print("⚠️   Missing 'certainty_percent' in output! Force-adding fallback.")
+            parsed["certainty_percent"] = 0
+        elif not isinstance(parsed["certainty_percent"], (int, float)):
+            parsed["certainty_percent"] = int(parsed["certainty_percent"])
+
+        # Clamp certainty to [0, 100]
+        pct = parsed["certainty_percent"]
+        parsed["certainty_percent"] = max(0, min(100, int(pct)))
+
+        # Enforce EDSS rules
+        if not parsed.get("klassifizierbar", False):
+            if "EDSS" in parsed:
+                del parsed["EDSS"]
+        else:
+            if "EDSS" not in parsed:
+                print("⚠️   'klassifizierbar' is true but EDSS missing — adding fallback.")
+                parsed["EDSS"] = 7.0
+
+        inference_time = time.time() - start_time
+
+        return {
+            "success": True,
+            "result": parsed,
+            "inference_time_sec": inference_time
+        }
+
+    except Exception as e:
+        print(f"❌ Inference error: {e}")
+        return {
+            "success": False,
+            "error": str(e),
+            "inference_time_sec": -1,
+            "result": None
+        }
+
+# === BUILD PATIENT TEXT ===
+def build_patient_text(row):
+    return (
+        str(row.get("T_Zusammenfassung", "")) + "\n" +
+        str(row.get("Diagnosen", "")) + "\n" +
+        str(row.get("T_KlinBef", "")) + "\n" +
+        str(row.get("T_Befunde", ""))
+    )
+
+# === MAIN LOOP (NEW: MULTI-ITERATION) ===
+if __name__ == "__main__":
+    # Load data ONCE (to avoid repeated I/O overhead)
+    df = pd.read_csv(INPUT_CSV, sep=';')
+    total_rows = len(df)
+    print(f"Loaded {total_rows} patient records.")
+
+    for iteration in range(1, NUM_ITERATIONS + 1):
+        print(f"\n{'='*60}")
+        print(f"🔄 ITERATION {iteration}/{NUM_ITERATIONS}")
+        print(f"{'='*60}")
+
+        iteration_results = []
+        start_iter = time.time()
+
+        for idx, row in df.iterrows():
+            print(f"\rRow {idx+1}/{total_rows} | Iter {iteration}", end='', flush=True)
+            try:
+                patient_text = build_patient_text(row)
+                result = run_inference(patient_text)
+
+                # Attach metadata
+                if result["success"]:
+                    res = result["result"].copy()  # avoid mutation
+                    res["iteration"] = iteration
+                    res["unique_id"] = row.get("unique_id", f"row_{idx}")
+                    res["MedDatum"] = row.get("MedDatum", None)
+                    result["result"] = res
+
+                else:
+                    result["iteration"] = iteration
+                    result["unique_id"] = row.get("unique_id", f"row_{idx}")
+                    result["MedDatum"] = row.get("MedDatum", None)
+
+                iteration_results.append(result)
+
+                if result["success"]:
+                    res = result["result"]
+                    edss = res.get("EDSS", "N/A") if res.get("klassifizierbar") else "N/A"
+                    print(f" ✅ EDSS={edss}, cert={res.get('certainty_percent', '?')}%")
+                else:
+                    print(f" ❌ {result.get('error', 'Unknown')}")
+
+            except Exception as e:
+                print(f"\n⚠️   Row {idx} failed: {e}")
+                iteration_results.append({
+                    "success": False,
+                    "error": str(e),
+                    "iteration": iteration,
+                    "unique_id": row.get("unique_id", f"row_{idx}"),
+                    "MedDatum": row.get("MedDatum", None),
+                    "result": None
+                })
+                if STOP_ON_FIRST_ERROR:
+                    break
+
+        # Save per-iteration results
+        timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
+        output_path = INPUT_CSV.replace(".csv", f"_results_iter_{iteration}_{timestamp}.json")
+        with open(output_path, 'w', encoding='utf-8') as f:
+            json.dump(iteration_results, f, indent=2, ensure_ascii=False)
+        print(f"\n✅ Iteration {iteration} complete. Saved to: {output_path}")
+
+        elapsed = time.time() - start_iter
+        print(f"⏱️   Iteration {iteration} took {elapsed:.1f}s ({elapsed/total_rows:.1f}s/row)")
+
+    print(f"\n🎉 All {NUM_ITERATIONS} iterations completed!")
+
+
+
+##

organize_project.sh

@@ -1,384 +0,0 @@
-#!/usr/bin/env bash
-
-set -euo pipefail
-
-# ============================================================
-# Organize Barcelona EDSS project safely
-# - Creates a timestamped backup first
-# - Creates a cleaner folder structure
-# - Moves files conservatively
-# - Does NOT delete anything
-# ============================================================
-
-PROJECT_ROOT="$(pwd)"
-TIMESTAMP="$(date +%Y%m%d_%H%M%S)"
-BACKUP_PARENT="${PROJECT_ROOT}/../Barcelona_backups"
-BACKUP_DIR="${BACKUP_PARENT}/Barcelona_backup_${TIMESTAMP}"
-
-echo "Project root: ${PROJECT_ROOT}"
-echo "Backup dir:   ${BACKUP_DIR}"
-echo
-
-# ------------------------------------------------------------
-# Safety checks
-# ------------------------------------------------------------
-
-if [ ! -f "${PROJECT_ROOT}/README.md" ]; then
-    echo "WARNING: README.md not found. Are you sure you are in the project root?"
-    echo "Current directory: ${PROJECT_ROOT}"
-    read -r -p "Continue anyway? [y/N] " answer
-    case "$answer" in
-        y|Y|yes|YES) ;;
-        *) echo "Aborted."; exit 1 ;;
-    esac
-fi
-
-if [ -d "${PROJECT_ROOT}/.git" ]; then
-    if ! git diff --quiet || ! git diff --cached --quiet; then
-        echo "ERROR: Git working tree is not clean."
-        echo "Please commit or stash changes before organizing."
-        exit 1
-    fi
-fi
-
-echo "This script will:"
-echo "1. Create a full backup."
-echo "2. Create organized folders."
-echo "3. Move files into data/, prompts/, scripts/, results/, archive/."
-echo "4. Keep your original files in the backup."
-echo
-read -r -p "Proceed? [y/N] " answer
-case "$answer" in
-    y|Y|yes|YES) ;;
-    *) echo "Aborted."; exit 1 ;;
-esac
-
-# ------------------------------------------------------------
-# Backup
-# ------------------------------------------------------------
-
-mkdir -p "${BACKUP_PARENT}"
-
-echo
-echo "Creating backup..."
-rsync -a \
-    --exclude "enarcelona/" \
-    --exclude "env/" \
-    --exclude ".venv/" \
-    --exclude "__pycache__/" \
-    "${PROJECT_ROOT}/" "${BACKUP_DIR}/"
-
-echo "Backup created at:"
-echo "${BACKUP_DIR}"
-
-# ------------------------------------------------------------
-# Create target structure
-# ------------------------------------------------------------
-
-echo
-echo "Creating new directory structure..."
-
-mkdir -p \
-    data/raw \
-    data/processed \
-    data/ground_truth \
-    data/external \
-    prompts \
-    scripts \
-    results/benchmark_runs \
-    results/final_results/model_outputs \
-    results/figures \
-    results/tables \
-    results/logs \
-    manuscript/figures \
-    manuscript/tables \
-    archive/old_scripts \
-    archive/old_results \
-    archive/tmp \
-    archive/old_data \
-    archive/old_project_files
-
-# ------------------------------------------------------------
-# Helper move functions
-# ------------------------------------------------------------
-
-move_if_exists() {
-    src="$1"
-    dest="$2"
-
-    if [ -e "$src" ]; then
-        mkdir -p "$(dirname "$dest")"
-
-        if [ -e "$dest" ]; then
-            echo "SKIP: destination exists: $dest"
-        else
-            echo "MOVE: $src -> $dest"
-            mv "$src" "$dest"
-        fi
-    fi
-}
-
-move_glob_if_exists() {
-    pattern="$1"
-    dest_dir="$2"
-
-    mkdir -p "$dest_dir"
-
-    shopt -s nullglob
-    files=( $pattern )
-    shopt -u nullglob
-
-    for f in "${files[@]}"; do
-        base="$(basename "$f")"
-        dest="${dest_dir}/${base}"
-
-        if [ -e "$dest" ]; then
-            echo "SKIP: destination exists: $dest"
-        else
-            echo "MOVE: $f -> $dest"
-            mv "$f" "$dest"
-        fi
-    done
-}
-
-# ------------------------------------------------------------
-# Move prompts / attached instruction files
-# ------------------------------------------------------------
-
-echo
-echo "Moving prompt and instruction files..."
-
-move_if_exists "attach/Komplett.txt" "prompts/Komplett.txt"
-move_if_exists "attach/just_edss_schema.gbnf" "prompts/just_edss_schema.gbnf"
-move_if_exists "attach/just_edss_text.txt" "prompts/just_edss_text.txt"
-
-# Move leftover attach folder if empty or archive it
-if [ -d "attach" ]; then
-    if [ -z "$(ls -A attach)" ]; then
-        rmdir attach
-    else
-        move_if_exists "attach" "archive/old_project_files/attach"
-    fi
-fi
-
-# ------------------------------------------------------------
-# Move important data files
-# ------------------------------------------------------------
-
-echo
-echo "Moving data files..."
-
-move_if_exists "Data/MS_Briefe_400_with_unique_id_SHA3_explore_cleaned.csv" \
-    "data/processed/MS_Briefe_400_with_unique_id_SHA3_explore_cleaned.csv"
-
-move_if_exists "Data/MS_Briefe_400_with_unique_id_SHA3_explore_cleaned_unique.csv" \
-    "data/processed/MS_Briefe_400_with_unique_id_SHA3_explore_cleaned_unique.csv"
-
-move_if_exists "Data/Join_edssandsub.tsv" \
-    "data/ground_truth/Join_edssandsub.tsv"
-
-move_if_exists "Data/GT_Numbers.csv" \
-    "data/ground_truth/GT_Numbers.csv"
-
-move_if_exists "Data/Annika1.csv" \
-    "data/ground_truth/Annika1.csv"
-
-move_if_exists "Data/comparison.tsv" \
-    "data/ground_truth/comparison.tsv"
-
-move_if_exists "Data/edss_distribution_summary.csv" \
-    "data/processed/edss_distribution_summary.csv"
-
-move_if_exists "Data/empirical_confidence_table.csv" \
-    "data/processed/empirical_confidence_table.csv"
-
-move_if_exists "Data/functional_system_colors.json" \
-    "data/processed/functional_system_colors.json"
-
-move_if_exists "Data/Test.csv" \
-    "archive/tmp/Test.csv"
-
-move_if_exists "Data/Hernan" \
-    "data/external/Hernan"
-
-move_if_exists "Data/iteration" \
-    "archive/old_data/iteration"
-
-# Old generated JSON/results from Data folder
-move_glob_if_exists "Data/*results*.json" "archive/old_results"
-move_glob_if_exists "Data/join_*.tsv" "archive/old_results"
-
-# Move remaining Data folder if anything left
-if [ -d "Data" ]; then
-    if [ -z "$(ls -A Data)" ]; then
-        rmdir Data
-    else
-        move_if_exists "Data" "archive/old_data/Data_remaining"
-    fi
-fi
-
-# ------------------------------------------------------------
-# Move benchmark results
-# ------------------------------------------------------------
-
-echo
-echo "Moving benchmark results..."
-
-if [ -d "results_edss_benchmark" ]; then
-    move_glob_if_exists "results_edss_benchmark/run_*" "results/benchmark_runs"
-
-    move_if_exists "results_edss_benchmark/endresults" \
-        "results/final_results/model_outputs"
-
-    move_if_exists "results_edss_benchmark/confusion_matrices" \
-        "results/figures/confusion_matrices"
-
-    if [ -z "$(ls -A results_edss_benchmark 2>/dev/null || true)" ]; then
-        rmdir results_edss_benchmark
-    else
-        move_if_exists "results_edss_benchmark" \
-            "archive/old_results/results_edss_benchmark_remaining"
-    fi
-fi
-
-# ------------------------------------------------------------
-# Move old/general results
-# ------------------------------------------------------------
-
-echo
-echo "Moving existing results files..."
-
-if [ -d "results" ]; then
-    # Figures
-    move_glob_if_exists "results/*.png" "results/figures"
-    move_glob_if_exists "results/*.PNG" "results/figures"
-    move_glob_if_exists "results/*.jpg" "results/figures"
-    move_glob_if_exists "results/*.jpeg" "results/figures"
-    move_glob_if_exists "results/*.svg" "results/figures"
-
-    # Tables
-    move_glob_if_exists "results/*.csv" "results/tables"
-    move_glob_if_exists "results/*.tsv" "results/tables"
-    move_glob_if_exists "results/*.xlsx" "results/tables"
-
-    # Subfolders that look like old results
-    move_if_exists "results/Jan_visual" "archive/old_results/Jan_visual"
-    move_if_exists "results/Lab_meeting" "archive/old_results/Lab_meeting"
-    move_if_exists "results/just_edss" "archive/old_results/just_edss"
-fi
-
-# Root-level result tables
-move_if_exists "edss_distribution_summary.csv" \
-    "results/tables/edss_distribution_summary.csv"
-
-# Logs
-move_if_exists "edss_benchmark_terminal.log" \
-    "results/logs/edss_benchmark_terminal.log"
-
-# ------------------------------------------------------------
-# Move scripts
-# ------------------------------------------------------------
-
-echo
-echo "Moving scripts..."
-
-move_if_exists "audit.py" "scripts/audit_outputs.py"
-move_if_exists "certainty.py" "scripts/analyze_certainty.py"
-move_if_exists "certainty_show.py" "scripts/certainty_show.py"
-move_if_exists "figure1.py" "scripts/figure1.py"
-move_if_exists "show_plots.py" "scripts/show_plots.py"
-
-move_if_exists "show_plots.py.orig" "archive/old_scripts/show_plots.py.orig"
-
-# Apps / old entry points
-move_if_exists "app.py" "archive/old_scripts/app.py"
-move_if_exists "total_app.py" "archive/old_scripts/total_app.py"
-
-# Existing project visuals folder
-move_if_exists "project/visuals" "results/figures/project_visuals"
-
-if [ -d "project" ]; then
-    if [ -z "$(ls -A project)" ]; then
-        rmdir project
-    else
-        move_if_exists "project" "archive/old_project_files/project"
-    fi
-fi
-
-# ------------------------------------------------------------
-# Environment folder
-# ------------------------------------------------------------
-
-echo
-echo "Handling virtual environment..."
-
-if [ -d "enarcelona" ]; then
-    echo "Leaving virtual environment in place: enarcelona/"
-    echo "It should remain ignored by .gitignore."
-fi
-
-# ------------------------------------------------------------
-# Create README notes
-# ------------------------------------------------------------
-
-echo
-echo "Writing organization notes..."
-
-cat > "PROJECT_STRUCTURE.md" <<'EOF'
-# Project Structure
-
-This project was reorganized into:
-
-- `data/`
-  - `raw/`: original raw data, if retained locally
-  - `processed/`: cleaned or derived input data
-  - `ground_truth/`: manually annotated reference data
-  - `external/`: externally provided data
-
-- `prompts/`
-  - EDSS instructions and prompt/schema assets
-
-- `scripts/`
-  - runnable analysis and plotting scripts
-
-- `results/`
-  - `benchmark_runs/`: full model benchmark runs
-  - `final_results/`: final selected model outputs
-  - `figures/`: generated figures
-  - `tables/`: generated tables
-  - `logs/`: terminal logs
-
-- `manuscript/`
-  - final figures and tables for paper/thesis writing
-
-- `archive/`
-  - old scripts, old results, temporary files, and unclear legacy files
-
-Important:
-The reorganization was performed after creating a full timestamped backup.
-EOF
-
-# ------------------------------------------------------------
-# Final checks
-# ------------------------------------------------------------
-
-echo
-echo "Organization complete."
-echo
-echo "Backup is here:"
-echo "${BACKUP_DIR}"
-echo
-echo "New top-level structure:"
-find . -maxdepth 2 -type d | sort
-echo
-
-if [ -d ".git" ]; then
-    echo "Git status:"
-    git status --short
-fi
-
-echo
-echo "Next recommended commands:"
-echo "  git status"
-echo "  git add ."
-echo "  git commit -m \"Reorganize project structure\""

prompts/Komplett.txt

@@ -1,481 +0,0 @@
-1 VISUAL OPTIC FUNCTIONS
-
-VISUAL ACUITY
-The visual acuity score is based on the line in the Snellen chart at 20 feet 5 meters
-for which the patient makes no more than one error using best available correction
-Alternatively best corrected near vision can be assessed but this should be noted and
-consistently performed during follow up examinations Switching from near to distance
-visual acuity measurements should be avoided in follow up examinations
-
-VISUAL FIELDS
-0 normal
-1 signs only deficits present only on formal confrontational testing
-2 moderate patient aware of deficit but incomplete hemianopsia on examination
-3 marked complete homonymous hemianopsia or equivalent
-
-SCOTOMA
-0 none
-1 small detectable only on formal confrontational testing
-2 large spontaneously reported by patient
-
-* DISC PALLOR
-0 not present
-1 present
-
-NOTE
-When determining the EDSS step the Visual FS score must be converted to a lower
-score as follows
-Visual FS Score 6 5 4 3 2 1
-Converted Visual FS Score 4 3 3 2 2 1
-
-FUNCTIONAL SYSTEM SCORE
-0 normal
-1 disc pallor and or small scotoma and or visual acuity corrected of worse eye less than 20 20 1.0 but better than 20 30 0.67
-2 worse eye with maximal visual acuity corrected of 20 30 to 20 59 0.67 – 0.34
-3 worse eye with large scotoma and or moderate decrease in fields and or maximal visual acuity corrected of 20 60 to 20 99 0.33 – 0.21
-4 worse eye with marked decrease of fields and or maximal visual acuity corrected of 20 100 to 20 200 0.2 – 0.1 grade 3 plus maximal acuity of better eye of 20 60 0.33 or less
-5 worse eye with maximal visual acuity corrected less than 20 200 0.1 grade 4 plus maximal acuity of better eye of 20 60 0.33 or less
-6 grade 5 plus maximal acuity of better eye of 20 60 0.33 or less *  =  optional part of the examination
-
-
-### BRAINSTEM FUNCTIONS
-
-**DYSARTHRIA**
-- **0**: None
-- **1**: Signs only
-- **2**: Mild: Clinically detectable, patient is aware
-- **3**: Moderate: Obvious during conversation, impairs comprehension
-- **4**: Marked: Incomprehensible speech
-- **5**: Inability to speak
-
-**DYSPHAGIA**
-- **0**: None
-- **1**: Signs only
-- **2**: Mild: Difficulty with thin liquids
-- **3**: Moderate: Difficulty with liquids and solid food
-- **4**: Marked: Sustained difficulty, requires pureed diet
-- **5**: Inability to swallow
-
-**OTHER CRANIAL NERVE FUNCTIONS**
-- **0**: Normal
-- **1**: Signs only
-- **2**: Mild disability: Clinically detectable deficit, patient is usually aware
-- **3**: Moderate disability
-- **4**: Marked disability
-
-**EXTRAOCULAR MOVEMENTS (EOM) IMPAIRMENT**
-- **0**: None
-- **1**: Signs only: Subtle EOM weakness, no complaints of vision issues
-- **2**: Mild: Subtle EOM weakness or obvious incomplete paralysis not noticed by patient
-- **3**: Moderate: Obvious incomplete paralysis noticed by patient or complete loss in one direction
-- **4**: Marked: Complete loss in more than one direction
-
-**NYSTAGMUS**
-- **0**: None
-- **1**: Signs only or mild: Gaze-evoked nystagmus below moderate limits (equivalent to Brainstem FS score of 1)
-- **2**: Moderate: Sustained nystagmus on horizontal/vertical gaze at 30 degrees, patient may not notice
-- **3**: Severe: Nystagmus in primary position or coarse persistent nystagmus interfering with vision; complete internuclear ophthalmoplegia; oscillopsia
-
-**TRIGEMINAL DAMAGE**
-- **0**: None
-- **1**: Signs only
-- **2**: Mild: Clinically detectable numbness, patient is aware
-- **3**: Moderate: Impaired sharp/dull discrimination in one to three branches or trigeminal neuralgia (at least one recent attack)
-- **4**: Marked: Unable to discriminate between sharp/dull or complete loss of sensation in one or both nerves
-
-**FACIAL WEAKNESS**
-- **0**: None
-- **1**: Signs only
-- **2**: Mild: Clinically detectable weakness, patient is aware
-- **3**: Moderate: Incomplete facial palsy (e.g., eye closure requires patching, drooling)
-- **4**: Marked: Complete unilateral or bilateral facial palsy with lagophthalmus or difficulty with liquids
-
-**HEARING LOSS**
-- **0**: None
-- **1**: Signs only: Hears finger rub less on one/both sides, lateralized Weber test but no complaints
-- **2**: Mild: As in 1, aware of hearing problem
-- **3**: Moderate: Does not hear finger rub on one/both sides, misses several whispered numbers
-- **4**: Marked: Misses all or nearly all whispered numbers
-
-**FUNCTIONAL SYSTEM SCORE**
-- **0**: Normal
-- **1**: Signs only
-- **2**: Moderate nystagmus/EOM impairment/other mild disability
-- **3**: Severe nystagmus/marked EOM impairment/moderate other cranial nerve disability
-- **4**: Marked dysarthria/other marked disability
-- **5**: Inability to swallow or speak
-
-### PYRAMIDAL FUNCTIONS
-
-#### REFLEXES
-- **0**: Absent
-- **1**: Diminished
-- **2**: Normal
-- **3**: Exaggerated
-- **4**: Nonsustained clonus (a few beats of clonus)
-- **5**: Sustained clonus
-
-##### Cutaneous Reflexes
-- **0**: Normal
-- **1**: Weak
-- **2**: Absent
-
-###### Palmomental Reflex
-- **0**: Absent
-- **1**: Present
-
-###### Plantar Response
-- **0**: Flexor
-- **1**: Neutral or equivocal
-- **2**: Extensor
-
-#### LIMB STRENGTH
-The weakest muscle in each group defines the score for that muscle group. Optional functional tests (hopping on one foot and walking on heels/toes) are recommended for BMRC grades 3–5.
-
-##### BMRC Rating Scale
-- **0**: No muscle contraction detected
-- **1**: Visible contraction without visible joint movement
-- **2**: Visible movement only on the plane of gravity
-- **3**: Active movement against gravity, but not against resistance
-- **4**: Active movement against resistance, but not full strength
-- **5**: Normal strength
-
-#### FUNCTIONAL TESTS
-##### Pronator Drift (Upper Extremities)
-Pronation and downward drift:
-- **0**: None
-- **1**: Mild
-- **2**: Evident
-
-##### Position Test (Lower Extremities)
-Ask patient to lift both legs together, with legs fully extended at the knee. Sinking:
-- **0**: None
-- **1**: Mild
-- **2**: Evident
-- **3**: Able to lift only one leg at a time (grade from the horizontal position at the hip joints in degrees)
-- **4**: Unable to lift one leg at a time
-
-##### Walking on Heels/Toes
-- **0**: Normal
-- **1**: Impaired
-- **2**: Not possible
-
-##### Hopping on One Foot
-- **0**: Normal
-- **1**: 6–10 times
-- **2**: 1–5 times
-- **3**: Not possible
-
-#### LIMB SPASTICITY (AFTER RAPID FLEXION OF THE EXTREMITY)
-- **0**: None
-- **1**: Mild: barely increased muscle tone
-- **2**: Moderate: moderately increased muscle tone that can be overcome; full range of motion is possible
-- **3**: Severe: severely increased muscle tone that is extremely difficult to overcome; full range of motion is not possible
-- **4**: Contracted
-
-#### GAIT SPASTICITY
-- **0**: None
-- **1**: Barely perceptible
-- **2**: Evident: minor interference with function
-- **3**: Permanent shuffling: major interference with function
-
-#### OVERALL MOTOR PERFORMANCE
-- **0**: Normal
-- **1**: Abnormal weakness (as compared to peers) in performing more demanding tasks, e.g., walking longer distances; no reduction in limb strength on formal testing
-- **2**: Reduction in strength of individual muscle groups at confrontational testing
-
-#### FUNCTIONAL SYSTEM SCORE
-- **0**: Normal
-- **1**: Abnormal signs without disability
-- **2**: Minimal disability: patient complains of motor-fatigability or reduced performance in strenuous motor tasks (motor performance grade 1) and/or BMRC grade 4 in one or two muscle groups
-- **3**: Mild to moderate paraparesis or hemiparesis: usually BMRC grade 4 in more than two muscle groups; and/or BMRC grade 3 in one or two muscle groups (movements against gravity
-are possible); and/or severe monoparesis: BMRC grade 2 or less in one muscle group
-- **4**: Marked paraparesis or hemiparesis: usually BMRC grade 2 in two limbs or monoplegia with BMRC grade 0 or 1 in one limb; and/or moderate tetraparesis: BMRC grade 3 in three or more limbs
-- **5**: Paraplegia: BMRC grade 0 or 1 in all muscle groups of the lower limbs; and/or marked tetraparesis: BMRC grade 2 or less in three or more limbs; and/or hemiplegia
-- **6**: Tetraplegia: BMRC grade 0 or 1 in all muscle groups of the upper and lower limbs
-
-### CEREBELLAR FUNCTIONS
-
-#### HEAD TREMOR
-- **0**: none
-- **1**: mild
-- **2**: moderate
-- **3**: severe
-
-#### TRUNCAL ATAXIA
-- **0**: none
-- **1**: signs only
-- **2**: mild (swaying with eyes closed)
-- **3**: moderate (swaying with eyes open)
-- **4**: severe (unable to sit without assistance)
-
-#### LIMB ATAXIA (TREMOR / DYSMETRIA AND RAPID ALTERNATING MOVEMENTS)
-- **0**: none
-- **1**: signs only
-- **2**: mild (tremor or clumsy movements easily seen, minor interference with function)
-- **3**: moderate (tremor or clumsy movements interfere with function in all spheres)
-- **4**: severe (most functions are very difficult)
-
-#### TANDEM (STRAIGHT LINE) WALKING
-- **0**: normal
-- **1**: impaired
-- **2**: not possible
-
-#### GAIT ATAXIA
-- **0**: none
-- **1**: signs only
-- **2**: mild (problems with balance realized by patient and/or significant other)
-- **3**: moderate (abnormal balance with ordinary walking)
-- **4**: severe (unable to walk more than a few steps unassisted or requires a walking aid or assistance due to ataxia)
-
-#### ROMBERG TEST
-- **0**: normal
-- **1**: mild (mild instability with eyes closed)
-- **2**: moderate (not stable with eyes closed)
-- **3**: severe (not stable with eyes open)
-
-#### OTHER CEREBELLAR TESTS
-- **0**: normal
-- **1**: mild abnormality
-- **2**: moderate abnormality
-- **3**: severe abnormality
-
-**NOTE:**
-- The presence of severe gait and/or truncal ataxia alone (without severe ataxia in three or four limbs) results in a Cerebellar FS score of 3.
-- If weakness or sensory deficits interfere with the testing of ataxia, score the patient’s actual performance. Indicate the possible role of weakness by marking an "X" after the
-affected subsystems and Cerebellar FS score.
-
-#### FUNCTIONAL SYSTEM SCORE
-- **0**: normal
-- **1**: abnormal signs without disability
-- **2**: mild ataxia and/or moderate station ataxia (Romberg) and/or tandem walking not possible
-- **3**: moderate limb ataxia and/or moderate or severe gait/truncal ataxia
-- **4**: severe gait/truncal ataxia and severe ataxia in three or four limbs
-- **5**: unable to perform coordinated movements due to ataxia
-- **X**: pyramidal weakness (BMRC grade 3 or worse in limb strength) or sensory deficits interfere with cerebellar testing
-
-### SENSORY FUNCTIONS
-
-#### SUPERFICIAL SENSATION (LIGHT TOUCH AND PAIN)
-- **0**: normal
-- **1**: signs only (slightly diminished sensation on formal testing, patient not aware)
-- **2**: mild (patient aware of impaired light touch or pain but can discriminate sharp/dull)
-- **3**: moderate (impaired discrimination of sharp/dull)
-- **4**: marked (unable to discriminate between sharp/dull and/or unable to feel light touch)
-- **5**: complete loss (anesthesia)
-
-#### VIBRATION SENSE (AT THE MOST DISTAL JOINT)
-- **0**: normal
-- **1**: mild (graded tuning fork 5–7 of 8; detects more than 10 seconds but less than examiner)
-- **2**: moderate (graded tuning fork 1–4 of 8; detects between 2 and 10 sec.)
-- **3**: marked (complete loss of vibration sense)
-
-#### POSITION SENSE
-- **0**: normal
-- **1**: mild (1–2 incorrect responses, only distal joints affected)
-- **2**: moderate (misses many movements of fingers or toes; proximal joints affected)
-- **3**: marked (no perception of movement, astasia)
-
-* **LHERMITTE’S SIGN** (does not contribute to the Sensory FS score)
-  - **0**: negative
-  - **1**: positive
-
-* **PARAESTHESIAE (TINGLING)** (does not contribute to the Sensory FS score)
-  - **0**: none
-  - **1**: present
-
-#### FUNCTIONAL SYSTEM SCORE
-- **0**: normal
-- **1**: impaired superficial sensation in one or two limbs
-- **2**: mild impairment in more than two limbs, no major proprioceptive deficits
-- **3**: moderate impairment in more than two limbs with minor proprioceptive deficits
-- **4**: severe impairment in more than two limbs with significant proprioceptive deficits
-- **5**: loss of sensation in one or two limbs, significant proprioceptive deficits in most of the body below the head
-- **6**: essentially no sensation below the head
-
-### BOWEL AND BLADDER FUNCTIONS
-
-#### URINARY HESITANCY AND RETENTION
-- **0**: none
-- **1**: mild (no major impact on lifestyle)
-- **2**: moderate (urinary retention; frequent urinary tract infections)
-- **3**: severe (requires catheterization)
-- **4**: loss of function (overflow incontinence)
-
-#### URINARY URGENCY AND INCONTINENCE
-- **0**: none
-- **1**: mild (no major impact on lifestyle)
-- **2**: moderate (rare incontinence occurring no more than once a week; must wear pads)
-- **3**: severe (frequent incontinence occurring from several times a week to more than once a day; must wear urinal or pads)
-- **4**: loss of function (loss of bladder control)
-
-#### BLADDER CATHETERIZATION
-- **0**: none
-- **1**: intermittent self-catheterization
-- **2**: constant catheterization
-
-#### BOWEL DYSFUNCTION
-- **0**: none
-- **1**: mild (no incontinence, no major impact on lifestyle, mild constipation)
-- **2**: moderate (must wear pads or alter lifestyle to be near lavatory)
-- **3**: severe (in need of enemas or manual measures to evacuate bowels)
-- **4**: complete loss of function
-
-#### SEXUAL DYSFUNCTION
-**Male**
-- **0**: none
-- **1**: mild (difficulty maintaining erection during intercourse, but achieves erection and still has intercourse)
-- **2**: moderate (difficulty achieving erection, decreased libido, still has intercourse and reaches orgasm)
-- **3**: severe (marked decrease in libido, inability to achieve full erection, intercourse with difficulty, hypoorgasmia)
-- **4**: loss of function
-
-**Female**
-- **0**: none
-- **1**: mild (mild lack of lubrication, still sexually active and reaches orgasm)
-- **2**: moderate (dyspareunia, hypoorgasmia, decrease in sexual activity)
-- **3**: severe (marked decrease in sexual activity, anorgasmia)
-- **4**: loss of function
-
-**NOTE**
-When determining the EDSS step, the Bowel and Bladder FS score must be converted to a lower score as follows:
-- Bowel and Bladder FS Score: 6 → Converted Bowel and Bladder FS Score: 5
-- Bowel and Bladder FS Score: 5 → Converted Bowel and Bladder FS Score: 4
-- Bowel and Bladder FS Score: 4 → Converted Bowel and Bladder FS Score: 3
-- Bowel and Bladder FS Score: 3 → Converted Bowel and Bladder FS Score: 3
-- Bowel and Bladder FS Score: 2 → Converted Bowel and Bladder FS Score: 2
-- Bowel and Bladder FS Score: 1 → Converted Bowel and Bladder FS Score: 1
-
-Sexual dysfunction can be documented but generally does not impact the FS score due to assessment difficulties by examining physicians.
-
-### FUNCTIONAL SYSTEM SCORE
-- **0**: normal
-- **1**: mild urinary hesitancy, urgency, and/or constipation
-- **2**: moderate urinary hesitancy/retention and/or moderate urinary urgency/incontinence and/or moderate bowel dysfunction
-- **3**: frequent urinary incontinence or intermittent self-catheterization; needs enemas or manual measures to evacuate bowels
-- **4**: in need of almost constant catheterization
-- **5**: loss of bladder or bowel function (external or indwelling catheter)
-- **6**: loss of bowel and bladder function
-
-### CEREBRAL FUNCTIONS
-
-#### DEPRESSION AND EUPHORIA
-- **0**: none
-- **1**: present (Patient complains of depression or is considered depressed or euphoric by the investigator or significant other.)
-
-**Note**: Depression and Euphoria are documented on the scoring sheet but are not taken into consideration for FS and EDSS calculation.
-
-#### DECREASE IN MENTATION
-- **0**: none
-- **1**: signs only (not apparent to patient and/or significant other)
-- **2**: mild (Patient and/or significant other report mild changes in mentation. Examples include: impaired ability to follow a rapid course of association or survey complex matters;
-impaired judgment in certain demanding situations; capable of handling routine daily activities, but unable to tolerate additional stressors; intermittently symptomatic even with
-normal levels of stress; reduced performance; tendency toward negligence due to obliviousness or fatigue.)
-- **3**: moderate (Definite abnormalities on brief mental status testing, but still oriented to person, place, and time)
-- **4**: marked (Not oriented in one or two spheres (person, place, or time); marked effect on lifestyle)
-- **5**: dementia, confusion, and/or complete disorientation
-
-#### FATIGUE
-- **0**: none
-- **1**: mild (Does not usually interfere with daily activities)
-- **2**: moderate (Interferes but does not limit daily activities for more than 50%)
-- **3**: severe (Significant limitation in daily activities (> 50% reduction))
-
-**Note**: Because fatigue is difficult to evaluate objectively, in some studies it does not contribute to the Cerebral FS score or EDSS step. Please adhere to the study’s specific
-instructions.
-
-### FUNCTIONAL SYSTEM SCORE
-- **0**: normal
-- **1**: signs only in decrease in mentation; mild fatigue
-- **2**: mild decrease in mentation; moderate or severe fatigue
-- **3**: moderate decrease in mentation
-- **4**: marked decrease in mentation
-- **5**: dementia
-
-### AMBULATION
-
-**Unrestricted Ambulation**
-- The patient can walk a normal distance without assistance, comparable to healthy individuals of similar age and physical condition.
-- EDSS step can range from 0 to 5.0, depending on the Functional System (FS) scores.
-
-**Fully Ambulatory**
-- At least 500 meters of ambulation without assistance, but not unrestricted.
-- EDSS step can range from 2.0 to 5.0, depending on FS scores.
-- The Pyramidal and/or Cerebellar FS must be ≥ 2 to reflect this restriction in ambulation.
-
-**Ambulation < 500 Meters**
-- If the walking distance is less than 500 meters, the EDSS step must be ≥ 4.5, depending on the walking ranges provided by the ambulation score and combination of FS scores.
-- EDSS steps 5.5 to 8.0 are exclusively defined by the ability to ambulate and type of assistance required, or the ability to use a wheelchair.
-
-**Assistance Needed**
-- Definitions for EDSS steps 6.0 or 6.5 include both the type of assistance required when walking and the walking range.
-- Assistance by another person is equivalent to bilateral assistance.
-
-**Note:**
-- The ambulation score represents both the walking range and the type of assistance required.
-- This score replaces several checkboxes used previously on the scoring sheet but does not introduce new definitions.
-- Use of a wheelchair can now be scored on the scoring sheet.
-- Indicate the reported distance and time for the patient in the appropriate field on the scoring sheet, followed by the type of assistance and walking distance measured during assessment.
-
-
-### DISTANCE AND TIME REPORTED BY PATIENT
-
-**Maximal Unassisted Walking Distance**
-- Maximal unassisted walking distance reported by the patient (in meters) without rest or assistance.
-- Time required to walk the maximum distance according to the patient (in minutes).
-
-**Assistance**
-0. Without help or assistance (allowing use of an ankle-foot orthotic device, but no other assistive devices).
-1. Unilateral assistance: one stick/crutch/brace.
-2. Bilateral assistance: two sticks/crutches/braces or assistance by another person.
-3. Wheelchair.
-
-**Distance**
-- Measure the distance the patient can walk in meters.
-  - **Unassisted:** Observe walking for a minimum of 500 meters and measure time needed, if possible.
-  - **Assisted:** Observe walking with assistive devices or help from another person for a minimum of 130 meters, if possible.
-
----
-
-### AMBULATION SCORE
-
-0. Unrestricted
-1. Fully ambulatory
-2. ≥ 300 meters but < 500 meters, without help or assistance (EDSS 4.5 or 5.0)
-3. ≥ 200 meters but < 300 meters, without help or assistance (EDSS 5.0)
-4. ≥ 100 meters but < 200 meters, without help or assistance (EDSS 5.5)
-5. Walking range < 100 meters without assistance (EDSS 6.0)
-6. Unilateral assistance, ≥ 50 meters (EDSS 6.0)
-7. Bilateral assistance, ≥ 120 meters (EDSS 6.0)
-8. Unilateral assistance, < 50 meters (EDSS 6.5)
-9. Bilateral assistance, ≥ 5 meters but < 120 meters (EDSS 6.5)
-10. Uses wheelchair without help; unable to walk 5 meters even with aid, essentially restricted to wheelchair; wheels self and transfers alone; up and about in wheelchair for some 12 hours a day (EDSS 7.0)
-11. Uses wheelchair with help; unable to take more than a few steps; restricted to wheelchair; may need some help in transferring and wheeling self (EDSS 7.5)
-12. Essentially restricted to bed or chair or perambulated in wheelchair, but out of bed most of the day; retains many self-care functions; generally has effective use of arms (EDSS 8.0)
-
-Expanded Disability Status Scale (EDSS)
-
-0   - Normal neurological exam (all Functional Systems [FS] grade 0)
-1.0 - No disability, minimal signs in one FS (one FS grade 1)
-1.5 - No disability, minimal signs in more than one FS (more than one FS grade 1)
-2.0 - Minimal disability in one FS (one FS grade 2, others 0 or 1)
-2.5 - Minimal disability in two FS (two FS grades 2, others 0 or 1)
-3.0 - Moderate disability in one FS (one FS grade 3, others 0 or 1) though fully ambulatory;
-or mild disability in three or four FS (three/four FS grades 2, others 0 or 1) though fully ambulatory
-3.5 - Fully ambulatory but with moderate disability in one FS (one FS grade 3) and mild disability in one or two FS (one/two FS grade 2) and others 0 or 1;
-or fully ambulatory with two FS grades 3 (others 0 or 1);
-or fully ambulatory with five FS grades 2 (others 0 or 1)
-4.0 - Unable to walk > 25 feet without aid
-4.5 - Unable to walk > 100 feet without aid
-5.0 - Relies on a walking aid; unable to walk > 300 feet without resting
-5.5 - Relies on a walking aid; unable to walk > 200 feet without resting
-6.0 - Unable to walk more than 50 feet with or without aid; cannot stand unaided for five minutes
-6.5 - Unable to walk more than 10 feet with or without aid; cannot stand unaided for two minutes
-7.0 - Unable to walk 5 meters even with aid, essentially restricted to wheelchair; wheels self and transfers alone; up and about in wheelchair some 12 hours a day
-7.5 - Unable to take more than a few steps; restricted to wheelchair; may need some help in transferring and in wheeling self
-8.0 - Essentially restricted to bed or chair or perambulated in wheelchair, but out of bed most of the day; retains many self-care functions; generally has effective use of arms
-8.5 - Essentially restricted to bed much of the day; has some effective use of arm(s); retains some self-care functions
-9.0 - Helpless bed patient; can communicate and eat
-9.5 - Totally helpless bed patient; unable to communicate effectively or eat/swallow
-10  - Death due to MS
-
-

prompts/just_edss_schema.gbnf

@@ -1,11 +0,0 @@
-EDSS-kv ::= "\"EDSS\"" space ":" space number
-Reason ::= "\"" char{0,400} "\"" space
-Reason-kv ::= "\"Reason\"" space ":" space Reason
-boolean ::= ("true" | "false") space
-char ::= [^"\\\x7F\x00-\x1F] | [\\] (["\\bfnrt] | "u" [0-9a-fA-F]{4})
-decimal-part ::= [0-9]{1,16}
-integral-part ::= [0] | [1-9] [0-9]{0,15}
-nicht-klassifizierbar-kv ::= "\"nicht_klassifizierbar\"" space ":" space boolean
-number ::= ("-"? integral-part) ("." decimal-part)? ([eE] [-+]? integral-part)? space
-root ::= "{" space Reason-kv "," space nicht-klassifizierbar-kv ( "," space ( EDSS-kv ) )? "}" space
-space ::= | " " | "\n"{1,2} [ \t]{0,20}

prompts/just_edss_text.txt

@@ -1,25 +0,0 @@
-Expanded Disability Status Scale (EDSS)
-
-0   - Normal neurological exam (all Functional Systems [FS] grade 0)
-1.0 - No disability, minimal signs in one FS (one FS grade 1)
-1.5 - No disability, minimal signs in more than one FS (more than one FS grade 1)
-2.0 - Minimal disability in one FS (one FS grade 2, others 0 or 1)
-2.5 - Minimal disability in two FS (two FS grades 2, others 0 or 1)
-3.0 - Moderate disability in one FS (one FS grade 3, others 0 or 1) though fully ambulatory;
-or mild disability in three or four FS (three/four FS grades 2, others 0 or 1) though fully ambulatory
-3.5 - Fully ambulatory but with moderate disability in one FS (one FS grade 3) and mild disability in one or two FS (one/two FS grade 2) and others 0 or 1;
-or fully ambulatory with two FS grades 3 (others 0 or 1);
-or fully ambulatory with five FS grades 2 (others 0 or 1)
-4.0 - Unable to walk > 25 feet without aid
-4.5 - Unable to walk > 100 feet without aid
-5.0 - Relies on a walking aid; unable to walk > 300 feet without resting
-5.5 - Relies on a walking aid; unable to walk > 200 feet without resting
-6.0 - Unable to walk more than 50 feet with or without aid; cannot stand unaided for five minutes
-6.5 - Unable to walk more than 10 feet with or without aid; cannot stand unaided for two minutes
-7.0 - Unable to walk 5 meters even with aid, essentially restricted to wheelchair; wheels self and transfers alone; up and about in wheelchair some 12 hours a day
-7.5 - Unable to take more than a few steps; restricted to wheelchair; may need some help in transferring and in wheeling self
-8.0 - Essentially restricted to bed or chair or perambulated in wheelchair, but out of bed most of the day; retains many self-care functions; generally has effective use of arms
-8.5 - Essentially restricted to bed much of the day; has some effective use of arm(s); retains some self-care functions
-9.0 - Helpless bed patient; can communicate and eat
-9.5 - Totally helpless bed patient; unable to communicate effectively or eat/swallow
-10  - Death due to MS