For employers

Hire this AI Trainer

Sign in or create an account to invite AI Trainers to your job.

Invite to Job
J

Joakim Ngiciri

Senior Software Engineer in Contract Review, Compliance, and Legal Research

KENYA flag
Nairobi, Kenya
$40.00/hrExpertClickworkerAxiom AIAppen

Key Skills

Software

ClickworkerClickworker
Axiom AI
AppenAppen
Data Annotation TechData Annotation Tech
MercorMercor
RemotasksRemotasks
OneFormaOneForma
MindriftMindrift
Micro1
HumanaticHumanatic

Top Subject Matter

Legal Services & Contract Review
Regulatory Compliance & Risk Analysis
Legal Research & Document Analysis

Top Data Types

Computer Code ProgrammingComputer Code Programming
TextText
DocumentDocument

Top Task Types

Classification
Text Generation
Text Summarization
Question Answering
Evaluation Rating
Computer Programming Coding
Prompt Response Writing SFT
Point Key Point
Transcription

Freelancer Overview

Senior Software Engineer in Contract Review, Compliance, and Legal Research. Brings 12+ years of professional experience across complex professional workflows, research, and quality-focused execution. Education includes Certificate in Full-Stack Web Development, UCF Boot Camps (2024) and Bachelor of Science, University of Nairobi (2010).

ExpertEnglish

Labeling Experience

Deep Learning-Based Medical Image Classification for Disease Diagnosis Support

ImageClassification
Medical Image Classification System 1. Introduction Medical image classification has emerged as a significant application within the field of Artificial Intelligence, particularly in supporting clinical decision-making processes. This project aims to design and implement a system capable of automatically classifying medical images into predefined diagnostic categories using deep learning techniques. The system is intended to assist healthcare practitioners in the early detection and diagnosis of diseases, thereby improving efficiency and accuracy in medical analysis. 2. Objectives The primary objectives of this project are as follows: To develop a structured and accurately labeled dataset of medical images. To design and train a classification model based on Deep Learning techniques. To evaluate the performance of the model using standard classification metrics. To implement a prototype system that allows users to input medical images and receive predicted diagnostic outputs. 3. Scope of Data Collection and Labeling 3.1 Data Sources The dataset will be obtained from publicly available medical image repositories, such as chest X-ray and dermatological image datasets. Where applicable, additional data may be sourced from healthcare institutions, subject to ethical approval and data privacy regulations. 3.2 Data Labeling Images will be categorized into clinically relevant classes, such as normal and diseased, or into multiple disease categories depending on the dataset. The labeling process will emphasize: Accuracy and consistency Class balance to reduce bias Validation against credible medical sources 4. Data Preprocessing Prior to model training, the dataset will undergo preprocessing to enhance model performance. This includes: Resizing images to standardized dimensions Normalizing pixel intensity values Applying data augmentation techniques such as rotation, flipping, and scaling These steps are essential to improve generalization and reduce overfitting. 5. Model Development The classification system will be developed using Convolutional Neural Networks (CNNs), which are well-suited for image analysis tasks. Two approaches will be considered: Development of a custom CNN architecture Application of transfer learning using pre-trained models such as ResNet or VGG16 Implementation will be carried out using frameworks such as TensorFlow or PyTorch. 6. Evaluation Metrics The performance of the model will be assessed using the following metrics: Accuracy Precision Recall F1-score Confusion matrix Particular emphasis will be placed on recall, as minimizing false negatives is critical in medical diagnosis. 7. System Implementation A prototype system will be developed to demonstrate the functionality of the trained model. The system will include: An interface for uploading medical images A backend component for processing and classification Output displaying predicted class labels and confidence scores 8. Ethical Considerations This project will adhere to ethical standards in handling medical data. All datasets will be anonymized to protect patient identity, and the system will be positioned strictly as a decision-support tool rather than a replacement for professional medical judgment. 9. Limitations The project is subject to several limitations, including: Limited dataset size and diversity Potential labeling inaccuracies Model generalization constraints Lack of clinical validation in real-world environments 10. Expected Outcomes The project is expected to produce: A trained medical image classification model A labeled dataset suitable for machine learning tasks A functional prototype system A comprehensive evaluation report 11. Conclusion In summary, this project seeks to apply deep learning techniques to the classification of medical images, contributing to the growing integration of artificial intelligence in healthcare. The system aims to enhance diagnostic support while highlighting the importance of ethical considerations and model reliability in medical applications.

Medical Image Classification System 1. Introduction Medical image classification has emerged as a significant application within the field of Artificial Intelligence, particularly in supporting clinical decision-making processes. This project aims to design and implement a system capable of automatically classifying medical images into predefined diagnostic categories using deep learning techniques. The system is intended to assist healthcare practitioners in the early detection and diagnosis of diseases, thereby improving efficiency and accuracy in medical analysis. 2. Objectives The primary objectives of this project are as follows: To develop a structured and accurately labeled dataset of medical images. To design and train a classification model based on Deep Learning techniques. To evaluate the performance of the model using standard classification metrics. To implement a prototype system that allows users to input medical images and receive predicted diagnostic outputs. 3. Scope of Data Collection and Labeling 3.1 Data Sources The dataset will be obtained from publicly available medical image repositories, such as chest X-ray and dermatological image datasets. Where applicable, additional data may be sourced from healthcare institutions, subject to ethical approval and data privacy regulations. 3.2 Data Labeling Images will be categorized into clinically relevant classes, such as normal and diseased, or into multiple disease categories depending on the dataset. The labeling process will emphasize: Accuracy and consistency Class balance to reduce bias Validation against credible medical sources 4. Data Preprocessing Prior to model training, the dataset will undergo preprocessing to enhance model performance. This includes: Resizing images to standardized dimensions Normalizing pixel intensity values Applying data augmentation techniques such as rotation, flipping, and scaling These steps are essential to improve generalization and reduce overfitting. 5. Model Development The classification system will be developed using Convolutional Neural Networks (CNNs), which are well-suited for image analysis tasks. Two approaches will be considered: Development of a custom CNN architecture Application of transfer learning using pre-trained models such as ResNet or VGG16 Implementation will be carried out using frameworks such as TensorFlow or PyTorch. 6. Evaluation Metrics The performance of the model will be assessed using the following metrics: Accuracy Precision Recall F1-score Confusion matrix Particular emphasis will be placed on recall, as minimizing false negatives is critical in medical diagnosis. 7. System Implementation A prototype system will be developed to demonstrate the functionality of the trained model. The system will include: An interface for uploading medical images A backend component for processing and classification Output displaying predicted class labels and confidence scores 8. Ethical Considerations This project will adhere to ethical standards in handling medical data. All datasets will be anonymized to protect patient identity, and the system will be positioned strictly as a decision-support tool rather than a replacement for professional medical judgment. 9. Limitations The project is subject to several limitations, including: Limited dataset size and diversity Potential labeling inaccuracies Model generalization constraints Lack of clinical validation in real-world environments 10. Expected Outcomes The project is expected to produce: A trained medical image classification model A labeled dataset suitable for machine learning tasks A functional prototype system A comprehensive evaluation report 11. Conclusion In summary, this project seeks to apply deep learning techniques to the classification of medical images, contributing to the growing integration of artificial intelligence in healthcare. The system aims to enhance diagnostic support while highlighting the importance of ethical considerations and model reliability in medical applications.

2018 - 2025

Education

U

UCF Boot Camps

Certificate in Full-Stack Web Development, Full-Stack Web Development

Certificate in Full-Stack Web Development
2024 - 2024
U

University of Nairobi

Bachelor of Science, Computer Service

Bachelor of Science
2010 - 2010

Work History

H

Humana

Senior Software Engineer

Nairobi
2025 - 2026
R

Reality AI Lab

Senior Software Developer

Nairobi
2024 - 2025