Merge pull request #7 from niranjan93/calc_exp

KOs for Top 10 MDCalc formulas used by clinicians

KO/clinical_calculators/ascvd-2013.json

@@ -0,0 +1,26 @@
+{
+    "@context": "https://kgrid.org/koio/2.0/context",
+    "@id": "ascvd-2013",
+    "koio:kgrid": "2",
+    "@type": "koio:KnowledgeObject",
+    "dc:title": "ASCVD (Atherosclerotic Cardiovascular Disease) 2013 Risk Calculator from AHA/ACC",
+    "dc:identifier": "ark:ascvd",
+    "version": "v1.0",
+    "dc:description": "Estimates the patient's risk of developing their first myocardial infarction or stroke in the next 10 years",
+    "koio:contributors": "Kgrid Team",
+    "koio:keywords": [
+      "atherosclerotic cardiovascular disease",
+      "coronary artery disease",
+      "myocardial infarction",
+      "stroke",
+      "risk"
+    ],
+    "koio:hasKnowledge": {
+      "@id": "knowledge",
+      "@type": [
+        "Knowledge",
+        "Function"
+      ],
+      "implementedBy": "https://raw.githubusercontent.com/kgrid-lab/LLM_with_RAG_chatbot/main/KO/clinical_calculators/code/ascvd_2013.py"
+    }
+  }

KO/clinical_calculators/bmi.json

@@ -0,0 +1,26 @@
+{
+    "@context": "https://kgrid.org/koio/2.0/context",
+    "@id": "bmi",
+    "koio:kgrid": "2",
+    "@type": "koio:KnowledgeObject",
+    "dc:title": "Body Mass Index",
+    "dc:identifier": "ark:bmi",
+    "version": "v1.0",
+    "dc:description": "The Body Mass Index is computed using the patient's height in meters and weight in kilograms. It can be used to estimate whether a patient is underweight or overweight. 18.5 - 25 is the normal range.",
+    "koio:contributors": "Kgrid Team",
+    "koio:keywords": [
+      "body mass index",
+      "body weight",
+      "underweight",
+      "overweight",
+      "obesity"
+    ],
+    "koio:hasKnowledge": {
+      "@id": "knowledge",
+      "@type": [
+        "Knowledge",
+        "Function"
+      ],
+      "implementedBy": "https://raw.githubusercontent.com/kgrid-lab/LLM_with_RAG_chatbot/main/KO/clinical_calculators/code/bmi.py"
+    }
+  }

KO/clinical_calculators/bsa.json

@@ -0,0 +1,22 @@
+{
+    "@context": "https://kgrid.org/koio/2.0/context",
+    "@id": "bsa",
+    "koio:kgrid": "2",
+    "@type": "koio:KnowledgeObject",
+    "dc:title": "Body Surface Area",
+    "dc:identifier": "ark:bsa",
+    "version": "v1.0",
+    "dc:description": "Estimates the patient's body surface area in square meters using the Mosteller formula, given the patient's height in centimeters and weight in kilograms.",
+    "koio:contributors": "Kgrid Team",
+    "koio:keywords": [
+      "body surface area"
+    ],
+    "koio:hasKnowledge": {
+      "@id": "knowledge",
+      "@type": [
+        "Knowledge",
+        "Function"
+      ],
+      "implementedBy": "https://raw.githubusercontent.com/kgrid-lab/LLM_with_RAG_chatbot/main/KO/clinical_calculators/code/bsa.py"
+    }
+  }

KO/clinical_calculators/chadsvasc.json

@@ -0,0 +1,23 @@
+{
+    "@context": "https://kgrid.org/koio/2.0/context",
+    "@id": "chadsvasc",
+    "koio:kgrid": "2",
+    "@type": "koio:KnowledgeObject",
+    "dc:title": "CHA2DS2-VASc Score",
+    "dc:identifier": "ark:chadsvasc",
+    "version": "v1.0",
+    "dc:description": "A score representing the risk of stroke in a patient with atrial fibrillation. 0 is low-risk in males. 0 or 1 is low-risk in females. All other scores are higher risk.",
+    "koio:contributors": "Kgrid Team",
+    "koio:keywords": [
+      "creatinine clearance",
+      "renal function"
+    ],
+    "koio:hasKnowledge": {
+      "@id": "knowledge",
+      "@type": [
+        "Knowledge",
+        "Function"
+      ],
+      "implementedBy": "https://raw.githubusercontent.com/kgrid-lab/LLM_with_RAG_chatbot/main/KO/clinical_calculators/code/chadsvasc.py"
+    }
+  }

KO/clinical_calculators/ckd-epi-gfr-2021.json

@@ -0,0 +1,23 @@
+{
+    "@context": "https://kgrid.org/koio/2.0/context",
+    "@id": "ckd-epi-gfr-2021",
+    "koio:kgrid": "2",
+    "@type": "koio:KnowledgeObject",
+    "dc:title": "2021 CKD-EPI equation for Glomerular Filtration Rate (GFR)",
+    "dc:identifier": "ark:ckdepigfr",
+    "version": "v1.0",
+    "dc:description": "Steady-state estimate of glomerular filtration rate (GFR) using 2021 CKD-EPI equations using either creatinine alone or both creatinine and cystatin-C.",
+    "koio:contributors": "Kgrid Team",
+    "koio:keywords": [
+      "glomerular filtration rate",
+      "renal function"
+    ],
+    "koio:hasKnowledge": {
+      "@id": "knowledge",
+      "@type": [
+        "Knowledge",
+        "Function"
+      ],
+      "implementedBy": "https://raw.githubusercontent.com/kgrid-lab/LLM_with_RAG_chatbot/main/KO/clinical_calculators/code/ckd_epi_gfr_2021.py"
+    }
+  }

KO/clinical_calculators/cockroft-gault-cr-cl.json

@@ -0,0 +1,23 @@
+{
+    "@context": "https://kgrid.org/koio/2.0/context",
+    "@id": "cockroft-gault-cr-cl",
+    "koio:kgrid": "2",
+    "@type": "koio:KnowledgeObject",
+    "dc:title": "Cockroft-Gault Creatinine Clearance",
+    "dc:identifier": "ark:cgcc",
+    "version": "v1.0",
+    "dc:description": "Estimates creatinine clearance in adults using the Cockfort-Gault model.",
+    "koio:contributors": "Kgrid Team",
+    "koio:keywords": [
+      "creatinine clearance",
+      "renal function"
+    ],
+    "koio:hasKnowledge": {
+      "@id": "knowledge",
+      "@type": [
+        "Knowledge",
+        "Function"
+      ],
+      "implementedBy": "https://raw.githubusercontent.com/kgrid-lab/LLM_with_RAG_chatbot/main/KO/clinical_calculators/code/cockroft_gault_cr_cl.py"
+    }
+  }

KO/clinical_calculators/code/ascvd_2013.py

@@ -0,0 +1,120 @@
+from math import e
+from math import log
+
+def ascvd_2013(age: int, dm: bool, sex: str, smoker: bool, total_cholesterol: int, hdl: int, sbp: int, htn_tx: bool, race: str) -> float:
+    """
+    Parameters:
+    - age: The patient's age in years. This calculation is only valid for ages 40-75.
+    - dm: True if the patient has diabetes. False otherwise.
+    - sex: The patient's sex, either male or female.
+    - smoker: True if the patient smokes.
+    - total_cholesterol: The patient's total serum cholesterol in miligrams per deciliter.
+    - hdl: The patient's serum high-density lipoprotein cholesterol in miligrams per decileter.
+    - sbp: The patient's systolic blood pressure in mm Hg.
+    - htn_tx: True if the patient is being treated for hypertension. False otherwise.
+    - race: The patient's race, either white or african american.
+    Returns: The patient's risk of developing their first myocardial infarction or stroke in the next 10 years.
+    """
+    coefficients = {}
+    values = {}
+
+    if race == "african american":
+        if sex == "female":
+            coefficients["ln_age"] = 17.114
+            coefficients["ln_age_sq"] = 0
+            coefficients["ln_tot_col"] = 0.940
+            coefficients["ln_age_ln_tot_col"] = 0
+            coefficients["ln_hdlc"] = -18.920
+            coefficients["ln_age_ln_hdlc"] = 4.475
+            coefficients["ln_tx_sbp"] = 29.291
+            coefficients["ln_age_ln_tx_sbp"] = -6.432
+            coefficients["ln_utx_sbp"] = 27.820
+            coefficients["ln_age_ln_utx_sbp"] = -6.087
+            coefficients["smoker"] = 0.691
+            coefficients["ln_age_smoker"] = 0
+            coefficients["diabetes"] = 0.874
+            baseline_survival = 0.9533
+            mean_score = 86.61
+        else:
+            coefficients["ln_age"] = 2.469
+            coefficients["ln_age_sq"] = 0
+            coefficients["ln_tot_col"] = 0.302
+            coefficients["ln_age_ln_tot_col"] = 0
+            coefficients["ln_hdlc"] = -0.307
+            coefficients["ln_age_ln_hdlc"] = 0
+            coefficients["ln_tx_sbp"] = 1.916
+            coefficients["ln_age_ln_tx_sbp"] = 0
+            coefficients["ln_utx_sbp"] = 1.809
+            coefficients["ln_age_ln_utx_sbp"] = 0
+            coefficients["smoker"] = 0.549
+            coefficients["ln_age_smoker"] = 0
+            coefficients["diabetes"] = 0.645
+            baseline_survival = 0.8954
+            mean_score = 19.54
+    else:
+        if sex == "female":
+            coefficients["ln_age"] = -29.799
+            coefficients["ln_age_sq"] = 4.884
+            coefficients["ln_tot_col"] = 13.540
+            coefficients["ln_age_ln_tot_col"] = -3.114
+            coefficients["ln_hdlc"] = -13.578
+            coefficients["ln_age_ln_hdlc"] = 3.149
+            coefficients["ln_tx_sbp"] = 2.019
+            coefficients["ln_age_ln_tx_sbp"] = 0
+            coefficients["ln_utx_sbp"] = 1.957
+            coefficients["ln_age_ln_utx_sbp"] = 0
+            coefficients["smoker"] = 7.574
+            coefficients["ln_age_smoker"] = -1.665
+            coefficients["diabetes"] = 0.661
+            baseline_survival = 0.9665
+            mean_score = -29.18
+        else:
+            coefficients["ln_age"] = 12.344
+            coefficients["ln_age_sq"] = 0
+            coefficients["ln_tot_col"] = 11.853
+            coefficients["ln_age_ln_tot_col"] = -2.664
+            coefficients["ln_hdlc"] = -7.990
+            coefficients["ln_age_ln_hdlc"] = 1.769
+            coefficients["ln_tx_sbp"] = 1.797
+            coefficients["ln_age_ln_tx_sbp"] = 0
+            coefficients["ln_utx_sbp"] = 1.764
+            coefficients["ln_age_ln_utx_sbp"] = 0
+            coefficients["smoker"] = 7.837
+            coefficients["ln_age_smoker"] = -1.795
+            coefficients["diabetes"] = 0.658
+            baseline_survival = 0.9144
+            mean_score = 61.18
+
+    if htn_tx:
+        coefficients["ln_utx_sbp"] = 0
+        coefficients["ln_age_ln_utx_sbp"] = 0
+    else:
+        coefficients["ln_tx_sbp"] = 0
+        coefficients["ln_age_ln_tx_sbp"] = 0
+
+    values["ln_age"] = log(age)
+    values["ln_age_sq"] = pow(values["ln_age"], 2)
+    values["ln_tot_col"] = log(total_cholesterol)
+    values["ln_age_ln_tot_col"] = values["ln_age"] * values["ln_tot_col"]
+    values["ln_hdlc"] = log(hdl)
+    values["ln_age_ln_hdlc"] = values["ln_age"] * values["ln_hdlc"]
+    values["ln_tx_sbp"] = log(sbp)
+    values["ln_age_ln_tx_sbp"] = values["ln_age"] * values["ln_tx_sbp"]
+    values["ln_utx_sbp"] = log(sbp)
+    values["ln_age_ln_utx_sbp"] = values["ln_age"] * values["ln_utx_sbp"]
+
+    if smoker:
+        values["smoker"] = 1
+    else:
+        values["smoker"] = 0
+
+    values["ln_age_smoker"] = values["ln_age"] * values["smoker"]
+
+    if dm:
+        values["diabetes"] = 1
+    else:
+        values["diabetes"] = 0
+
+    score = sum((coefficients[term] * values[term] for term in coefficients.keys()))
+
+    return 1 - pow(baseline_survival, pow(e, score - mean_score))

KO/clinical_calculators/code/bmi.py

@@ -0,0 +1,8 @@
+def bmi(height: float, weight: float) -> float:
+    """
+    Paramters:
+    - height: The patient's height in meters.
+    - weight: The patient's weight in kilograms.
+    Returns: The Body Mass Index
+    """
+    return weight / pow(height, 2)

KO/clinical_calculators/code/bsa.py

@@ -0,0 +1,8 @@
+def bsa(height: float, weight: float) -> float:
+    """
+    Paramters:
+    - height: The patient's height in centimeters.
+    - weight: The patient's weight in kilograms.
+    Returns: The patient's body surface area in square meters.
+    """
+    return pow(weight * height, 0.5) / 60

KO/clinical_calculators/code/chadsvasc.py

@@ -0,0 +1,36 @@
+def chadsvasc(age: int, sex: str, chf: bool, htn: bool, stroke: bool, vasc: bool, dm: bool) -> int:
+    """
+    Parameters:
+    - age: The patient's age in years.
+    - sex: The patient's sex, male or female.
+    - chf: True if the patient has a history of heart failure. False otherwise.
+    - htn: True if the patient has a history of hypertension. False otherwise.
+    - stroke: True if the patient has a history of stroke, TIA, or thromboembolism. False otherwise.
+    - vasc: True if the patient has a history of vascular disease, including prior myocardial infarction, peripheral arterial disease, or aortic plaque. False otherwise.
+    - dm: True if the patient has a history of diabetes. False otherwise.
+    Returns: A score representing the risk of stroke in a patient with atrial fibrillation. 0 is low-risk in males. 0 or 1 is low-risk in females. All other scores are higher risk.
+    """
+    score = 0
+
+    if age >= 65:
+        if age >= 75:
+            score += 2
+        else:
+            score += 1
+
+    if sex == "female":
+        score += 1
+
+    if htn:
+        score += 1
+
+    if stroke:
+        score += 2
+
+    if vasc:
+        score += 1
+
+    if dm:
+        score +=1
+
+    return score

KO/clinical_calculators/code/ckd_epi_gfr_2021.py

@@ -0,0 +1,30 @@
+def ckd_epi_gfr_2021(sex: str, age: int, creatinine: float, cystatinc=None) -> float:
+    """
+    Parameters:
+    - sex (str): Sex of the patient, male or female.
+    - age (int): Age of the patient in years.
+    - creatinine (float): Serum creatinine concentration for the patient, in milligrams per deciliter.
+    - cystatinc (float): [optional] Serum cystatin-c concentration for the patient, in milligrams per liter.
+    """
+    if cystatinc is None:
+        if sex == "female":
+            mult = 1.012
+            k = 0.7
+            alpha = -0.241
+        else:
+            mult = 1
+            k = 0.9
+            alpha = -0.302
+
+        return 142 * pow(min(creatinine / k, 1), alpha) * pow(max(creatinine / k, 1), -1.2) * pow(0.9938, age) * mult
+    else:
+        if sex == "female":
+            mult = 0.963
+            k = 0.7
+            alpha = -0.219
+        else:
+            mult = 1
+            k = 0.9
+            alpha = -0.144
+
+        return 135 * pow(min(creatinine / k, 1), alpha) * pow(max(creatinine / k, 1), -0.544) * pow(min(cystatinc / 0.8, 1), -0.323) * pow(max(cystatinc / 0.8, 1), -0.778) * pow(0.9961, age) * mult

KO/clinical_calculators/code/cockroft_gault_cr_cl.py

@@ -0,0 +1,14 @@
+def cockroft_gault_cr_cl(sex: str, age: int, weight: float, creatinine: float) -> float:
+    """
+    Parameters:
+    - sex (str): Sex of the patient, male or female.
+    - age (int): Age of the patient in years.
+    - weight (float): Weight of the patient in kilograms.
+    - creatinine (float): Serum creatinine concentration for the patient, in milligrams per deciliter.
+    """
+    if sex == "female":
+        mult = 0.85
+    else:
+        mult = 1
+
+    return (((140 - age) * weight) / (72 * creatinine)) * mult

KO/clinical_calculators/code/corr_ca_alb.py

@@ -0,0 +1,9 @@
+def corr_ca_alb(ca: float, albumin: float, nl_alb=4.0) -> float:
+    """
+    Parameters:
+    - ca: The patient's serum calcium level in miligrams per deciliter.
+    - albumin: The patient's serum albumin level in grams per deciliter.
+    - nl_alb: The normal serum albumin level. Optional. The default is 4 mg/dL.
+    Returns: The patient's effective serum calcium level, corrected for their albumin level.
+    """
+    return ca + ((nl_alb - albumin) * 0.8)