A new breed of next generation predictive/classifier AI-powered service that helps medics, healthcare practitioners (HCPs) or medical decision experts to make scientific decisions by performing binary or multi-class predictions and classifications for example or sanity checking their existing assessments, binary or multi-class predictions and classifications. Some examples:
1. Breast cancer prediction (malignant or benign classification): Having extracted and recorded data features by analysing hundreds of past breast tissues (in line with Wisconsin cancer data) including mean of distances from centre to points on the perimeter, standard deviation of grey-scale values, mean size of the core tumour, mean of local variation in radius lengths, mean compactness, mean concavity, mean of severity of concave portions of the contour, mean for number of concave portions of the contour, mean symmetry, mean fractal dimension, standard error for the mean of distances from centre to points on the perimeter, standard error for standard deviation of grey-scale values, standard error perimeter, standard error of area, standard error for local variation in radius lengths, standard error for compactness, standard error for severity of concave portions of the contour, standard error for number of concave portions of the contour, standard error for symmetry, standard error for "coastline approximation", "worst" or largest mean value for mean of distances from centre to points on the perimeter, "worst" or largest mean value for standard deviation of grey-scale values, "worst" or largest mean value for local variation in radius lengths, "worst" or largest mean value for severity of concave portions of the contour, "worst" or largest mean value for number of concave portions of the contour, "worst" or largest mean value for "coastline approximation" and any other relevant data features and having verified and labelled those past samples correctly as (M = malignant, B = benign, the labels and targets of classification), decide whether a new tissue is malignant or not (and with what probability and certainty), provided that the data features (described above) are available for the new sample.
2. Predicting future obesity in youngsters: Butterfly AI can help health authorities to predict the future obesity cases among younger population across different regions and districts. Having recorded the past data features of thousands of youngsters including the amount of food consumed above the recommended level per age, average number of sugary drinks per day, the number of minutes of moderate-intensive weekly aerobatic activity, The weight class of parent 1 (Insufficient Weight, Normal Weight, Overweight Level I, Overweight Level II, Obesity Type I, Obesity Type II and Obesity Type III), The weight of class parent 2 (Insufficient Weight, Normal Weight, Overweight Level I, Overweight Level II, Obesity Type I, Obesity Type II and Obesity Type III), The weight class of older sibling (Insufficient Weight, Normal Weight, Overweight Level I, Overweight Level II, Obesity Type I, Obesity Type II and Obesity Type III), presence of genetic conditions that can cause obesity such as Prader-Willi syndrome in family (Yes or No), Having underlying medical conditions that may contribute to weight gain including an underactive thyroid gland (hypothyroidism) or hormones Cushing's syndrome (Yes or NO), Number of obese people living within 5 km radius of young person’s home, average income of people living within 5 km radius of youngster’s home, average number of calories per day, average number of eating out at restaurants per week, average number of fresh vegetables consumed per day, average number of processed or fast food consumed per day, amount of sugar per pay, number of main meals, consumption of food between meals, consumption of water daily, daily consumption of alcohol, smoking, frequency of physical activity, daily screen time (TV, movies, etc.), type of transportation used, gender, age, height and weight, and having verified and labelled those past study cases as someone who became obese or not later in life, predict whether a youngster is going to become obese in future or not.
3. Patient survival prediction: Knowing the data feature of patients including age, BMI, elective surgery, ethnicity, gender, height, icu admit source, ICU stay type, ICU type, weight, diagnosis, post operative, heart rate, temperatures, ventilated, cirrhosis, diabetes mellitus, hepatic failure, immune suppression, leukemia, lymphoma, solid tumor with metastasis, etc. and knowing whether each one of those past patients died or survived hospital treatment or surgery operation, predict whether a new patient will also die or survive the hospital treatment or surgery operation.
4. Migraine classification: Knowing the data feature of past patients’ migraine including Age, Duration, Frequency, Location, Character, Intensity, Nausea, Vomit, Phonophobia, Photophobia, Visual, Sensory, Dysphasia, Dysarthria, Vertigo, Tinnitus, Hypoacusis, Diplopia, Defect, Ataxia, Conscience, Paresthesia, DPF, etc. and knowing exactly what type of migraine those past patients had (Typical aura with migraine, Migraine without aura, Basilar-type aura, Sporadic hemiplegic migraine or Familial hemiplegic migraine, etc.), classify what type of migraine a new patient has (Typical aura with migraine, Migraine without aura, Basilar-type aura, Sporadic hemiplegic migraine or Familial hemiplegic migraine, etc.).
5. Prediction of Hypothyroid: Knowing the data feature of patients including age, sex, on thyroxine, query on thyroxine, on antithyroid medication, sick, pregnant, thyroid surgery, I131 treatment, query hypothyroid, query hyperthyroid, lithium, goitre, tumour, hypopituitary, psych, TSH measured, TSH, T3 measured, T3, TT4 measured, TT4, T4U measured, T4U, FTI measured, FTI, TBG measured, TBG, referral source, etc. and knowing whether each one of those past patients have had hypothyroid or not, predict whether a new patient will also develop hypothyroid or not.
and any other binary or multi-class health-related medical classification assessment or predictive use cases.
Log in/Register Request a Demo Instruction Manual
