03.Computer Vision01.Image detection and classification - sporedata/researchdesigneR GitHub Wiki

1. Use cases: in which situations should I use this method?

To establish diagnosis of a given condition based on medical images - see Development and Validation of a Deep Learning Algorithm for Detection of Diabetic Retinopathy in Retinal Fundus Photographs
To establish staging of a given condition based on medical images - see Automated Grading of Age-Related Macular Degeneration From Color Fundus Images Using Deep Convolutional Neural Networks
To establish prognosis based on exams with a graphical output such as EKG - see An artificial intelligence-enabled ECG algorithm for the identification of patients with atrial fibrillation during sinus rhythm: a retrospective analysis of outcome prediction
To establish prognosis of a given condition - see Predicting cancer outcomes from histology and genomics using convolutional networks

2. Input: what kind of data does the method require?

Medical images such as radiographs, MRI and CT exams
Exams with a graphical output such as EKG

3. Algorithm: how does the method work?

Model mechanics

Describing in words

One of the most popular algorithms for image detection and classification is through the use of neural networks (the population convolutional networks being one of its subtypes). Neural networks have the following stages:

Forward propagation
Total error calculation
Gradient (derivative) calculation
Gradient checking
Updating weights

Hyperparameters are characteristics of the model that can be changed by the modeler to improve its performance, and include the number layers, nodes, learning rate, weight values, bias or offset value (a constant representing a node connected to all other nodes within the same layer, allowing the activation function to be shifted) hidden layers (determine deep learning) among many others.

Describing in images

Describing with code

Breaking down equations

Reporting guidelines

1.Recent Advances on Spectral–Spatial Hyperspectral Image Classification: An Overview and New Guidelines 2.Advanced Assay Development Guidelines for Image-Based High Content Screening and Analysis

Data science packages

Suggested companion methods

Learning materials

Books *
Articles
- On deep learning for medical image analysis [1].
- Using free-response receiver operating characteristic curves to assess the accuracy of machine diagnosis of cancer [2].

4. Output: how do I interpret this method's results?

Mock conclusions or most frequent format for conclusions reached at the end of a typical analysis.

Tables, plots, and their interpretation

5. SporeData-specific

Templates

Image recognition

Data science functions

References

[1] Carin L, Pencina MJ. On deep learning for medical image analysis . Jama. 2018 Sep 18;320(11):1192-3.

[2] Moskowitz CS. Moskowitz CS. Using free-response receiver operating characteristic curves to assess the accuracy of machine diagnosis of cancer. Jama. 2017 Dec 12;318(22):2250-1.cancer. Jama. 2017 Dec 12;318(22):2250-1.