Benchmarking - Malaria

Organized by herilalaina - Current server time: Oct. 27, 2020, 12:23 p.m. UTC

Previous

Microscopy Dataset
April 19, 2020, midnight UTC

Current

Upload Dataset
March 10, 2020, 1:43 a.m. UTC

End

Competition Ends
Never

Benchmarking Malaria Detection Algorithms

Malaria is one of the top 10 causes of death in the sub Saharan Africa. According to the World Health Organization report of 2016, of the 438,000 Malaria deaths registered, an estimated 92% of all Malaria cases resulted in deaths, two thirds of which occurred among children under five years of age. Referring also to records from WHO report of 2015, Malaria accounted for 480,000 deaths, 90% of which were from Africa, 7% from S.E Asia and 2% from Eastern Mediterranean region. Although there were fewer Malaria cases in 2017 than in 2010 according to WHO report of 2017, data for the period 2015–2017 highlighted that no significant progress in reducing global Malaria cases was made in this timeframe. Malaria is thus of major concern to public health. who_itu_malaria

Benchmark

Benchmarking of “AI-based detection of Malaria” aims at considering many aspects that concern the entire process of detection. These span from the data collection pipeline, the data, the analysis engine, the output, the evaluation measures, the test data set and the framework design. The topic will in future benchmark on the benefits from multiple data sets whose factors could potentially improve the detective index of the AI based detection model.

How to participate

Prerequisites: Install Anaconda Python 3.6.6, Tensorflow (2.1.0), opencv-python (4.0.1), scikit-image (0.15.0) Download the starting kit. Usage: - modify sample_code_submission/model.py to provide a better model - zip the contents of sample_code_submission (without the directory, but with metadata)

Evaluation

This benchmark only allow code submission. Metrics are:

  • ROC AUC
  • Precision
  • Recall
  • MAP

Rules

You may submit 5 submissions every day and 100 in total.

Download Size (mb) Phase
Starting Kit 0.446 #1 Microscopy Dataset
Input Data 91.072 #1 Microscopy Dataset
Starting Kit 0.815 #2 Blood Smear
Public Data 104.230 #2 Blood Smear
Public Data 0.779 #3 Upload Dataset

Upload Your Dataset

An example of valid submission is given in Public Data section. Each submission must contains 9 files:

  • BASENAME_feat.name: name of each feature
  • BASENAME_label.name: list of classes
  • BASENAME_private.info: Private information to be shared for organizers only
  • BASENAME_public.info: information to be shared for all participants (description of data, source, ...)
  • BASENAME_train.data: train data (1 row = 1 example)
  • BASENAME_train.solution: true solution for BASENAME_train.data
  • BASENAME_test.data
  • BASENAME_test.solution
  • metadata: empty file (mandatory for each Codalab Submission)

BASENAME will be used as identifier of your dataset. Please validate your dataset before submission. Validation scripts are available at https://github.com/herilalaina/benchmarking_malaria/tree/master/upload_dataset

Microscopy Dataset

Start: April 19, 2020, midnight

Description: Benchmark on Microscopy Dataset.

Blood Smear

Start: April 26, 2020, 1:23 a.m.

Description: Benchmark on Blood Smear Dataset.

Upload Dataset

Start: March 10, 2020, 1:43 a.m.

Description: Upload Dataset

Competition Ends

Never

You must be logged in to participate in competitions.

Sign In