Working with clinical notes

import re

note = """HISTORY: 58M with T2DM x 10 yrs, HTN. c/o tingling in both feet x 3 mo. Denies chest pain. No SOB.
MEDICATIONS: Metformin 1 g BD. Amlodipine 5 mg OD.
EXAMINATION: Reduced monofilament sensation both feet. No ulcers. Pedal pulses present.
ASSESSMENT: Probable diabetic peripheral neuropathy. HbA1c 8.4%.
PLAN: Refer for DPN testing. Foot care advice. Review in 3 mo."""

def split_sections(text):
    """Split a note on headings written in capitals followed by a colon."""
    parts = re.split(r"^([A-Z][A-Z ]+):\s*", text, flags=re.MULTILINE)[1:]
    return {parts[i].title(): parts[i + 1].strip() for i in range(0, len(parts), 2)}

for name, body in split_sections(note).items():
    print(f"[{name}] {body}")

# [History] 58M with T2DM x 10 yrs, HTN. c/o tingling in both feet x 3 mo. Denies chest pain. No SOB.
# [Medications] Metformin 1 g BD. Amlodipine 5 mg OD.
# [Examination] Reduced monofilament sensation both feet. No ulcers. Pedal pulses present.
# [Assessment] Probable diabetic peripheral neuropathy. HbA1c 8.4%.
# [Plan] Refer for DPN testing. Foot care advice. Review in 3 mo.

ABBREVIATIONS = {
    "T2DM": "type 2 diabetes mellitus", "HTN": "hypertension", "c/o": "complains of",
    "SOB": "shortness of breath", "BD": "twice daily", "OD": "once daily",
    "mo": "months", "yrs": "years", "DPN": "diabetic peripheral neuropathy",
}
def expand(text):
    for short, long in ABBREVIATIONS.items():
        text = re.sub(rf"(?<![\w/]){re.escape(short)}(?![\w/])", long, text)
    return text

print(expand(split_sections(note)["History"]))

# 58M with type 2 diabetes mellitus x 10 years, hypertension. complains of tingling in both feet x 3 months. Denies chest pain. No shortness of breath.

NEGATION_CUES = r"\b(no|denies|denied|without|negative for|absent)\b"
FINDINGS = ["chest pain", "shortness of breath", "ulcers", "tingling", "pedal pulses",
            "reduced monofilament sensation"]

def findings_with_status(text):
    """NegEx-style: a finding is negated if a negation cue appears earlier in the same sentence."""
    results = []
    for sentence in re.split(r"(?<=[.!?])\s+", text):
        s = sentence.lower()
        for f in FINDINGS:
            pos = s.find(f)
            if pos >= 0:
                negated = re.search(NEGATION_CUES, s[:pos]) is not None
                results.append((f, "absent" if negated else "present"))
    return results

for finding, status in findings_with_status(expand(note)):
    print(f"{status:<8} {finding}")

# present  tingling
# absent   chest pain
# absent   shortness of breath
# present  reduced monofilament sensation
# absent   ulcers
# present  pedal pulses

import difflib

visit_1 = "Tingling in both feet for 3 months. No ulcers. Metformin 1 g twice daily. Plan: refer for neuropathy testing."
visit_2 = visit_1 + " Neuropathy test done: reduced sural nerve response."
visit_3 = visit_2                                           # copied forward unchanged

for name, a, b in [("visit 2 vs 1", visit_1, visit_2), ("visit 3 vs 2", visit_2, visit_3)]:
    print(f"{name}: {difflib.SequenceMatcher(None, a, b).ratio():.0%} identical")

# visit 2 vs 1: 81% identical
# visit 3 vs 2: 100% identical