Introduction

It is estimated that 48.5 million couples are suffering from infertility worldwide, where 20-30% of cases originate from males, and the market is expected to expand. The global cost of male infertility is estimated at $301.5 million [1]. Computer-aided semen analysis (CASA) is a term that refers to automatic or semi-automatic semen analysis techniques. It involves automated analysis of the ejaculate using equipment employing imaging technology. In 2020, infertility in Algeria affected approximately 15% of couples. The Ministry of Health has quantified this rate to 300,000 infertile couples [2,3,4]. CASA is a method based on image analysis of video sequences. In recent years, semen analysis has become increasingly important in the treatment of fertility problems. Traditionally, sperm quality is assessed by an expert using a microscope to examine the samples. CASA Systems have been developed to help these experts measure the factors that impact sperm quality, such as semen volume, total sperm count, concentration, vitality, motility, or morphology, for the purpose of optimal sperm selection for fertility treatments [5,6].

1. Development Framework and Objectives for Proposed CASA System

1.1. Problematic / proposed solution

1.1.1 Problematic

The commercial systems need to be developed and understood on how to get the job done, i.e., how to develop static and dynamic image processing algorithms on a high-performance hardware platform. Closed (hidden) commercial systems need to develop their own system. Understand the system. How to get the job done: developing static and dynamic image processing algorithms on a high-performance hardware platform? Design: Actually, what are the alternatives and constraints to move forward in the project? The CASA, hardware & software, offers an easy-to-move, small, high-performance, affordable, reliable, and easy-to-use system. Features of an on-board electronic system, with the possibility of finding a compromise between the material and the concept of co-design. Thus, our system is based on Raspberry PI 4 Reinforced by remote processing servers (Hardware resource): Cloud, programming language: PYTHON (open source) free (software), and Matlab easier (programs exist) but with an expensive license (not practical).

1.1.2. Proposed Solution: Proposed CASA

The basic idea operates on the same principle and leverages advancements in computer and electronic technology:

1. Replacing the human eye with a currently highly sophisticated camera (CMOS).

2. Image processing (utilizing very advanced tools and programs) employing Artificial Intelligence (Machine Learning) within computer science.

3. Manual calculations by a high-performance processing unit (currently highly developed in electronics), offering several OPTIONS (GPU, CPU, etc.).

4. Device selection (Inputs, Outputs): Acquisition and Display (Electronic), such as a Touch Screen, for example.

5. Connectivity to the internet, employing IOT and CLOUD technologies.

1.2. Objectives

1.2.1. Main objectives

The primary aim is to fabricate a semi-automatic device capable of interpreting slides from various tests prepared by the technician. It proposes a semi-automatic CASA system that can be utilized in:

• Modern reproductive medicine.

• Utilizing computer image processing technology.

• Leveraging advanced artificial intelligence technology. This can be employed in:

• Automatic sperm identification.

• Automatic sperm movement tracking.

1.2.2. Secondary Objectives

• Patent filing.

• Developing a system capable of rectifying errors in both the technician's readings (biological intelligence) and those made by the automatic CASA (Artificial Intelligence, AI).

2. Materials & Methods

2.1. List of materials (minimum requirement)

• Raspberry Pi board. • A micro-SD card of 16 GB or more. • A 5 VDC micro-USB power supply.

• A micro-SD card reader.

• A keyboard and mouse.

• An HDMI cable (micro-HDMI for the Pi 4B). • A display with an HDMI input.

2.2. Technology Description

It is a Single-board computer booting from a micro-SD card and running on Linux OS or Windows 10 IoT:

• Raspberry Pi 4 Model B 4 GB ARM-Cortex-A72 4 x 1.50 GHz.

• 4 GB RAM, 4 x USB, 2 x Micro HDMI (costing 59 €).

• Four 64-bit cores supporting dual-screen functionality.

• Connectivity: Gigabit Ethernet (RJ45), WiFi 802.11b/g/n/ac 5, Bluetooth 5.0.

• Storage: micro-SD slot available.

• Dimensions: 10 x 7 x 3 cm, Weight: 50 grams.

Figure 1. The Raspberry PI 4: a PC-on-board (pictured).

Figure 2. Prototype of the proposed system

Conclusion

In our study, we have embarked on the development of an innovative automatic semen analysis system, which harnesses advanced image processing techniques and artificial intelligence (AI), implemented on an accessible, embedded system like Raspberry Pi. This state-of-the-art approach is designed to significantly enhance the accuracy and efficiency of sperm analysis, offering substantial improvements in the realms of reproductive diagnostics and male fertility treatments. By integrating AI into the semen analysis process, we anticipate a substantial improvement in diagnostic accuracy and patient care. This technology is poised to reduce processing times significantly, making these vital analyses more accessible, particularly in regions with limited resources. Furthermore, the system's capability to store and review patient data paves the way for highly personalized fertility tracking.

However, this innovative approach is not without its challenges. Key among these are ensuring robust data protection, accurately interpreting results generated by AI, and seamlessly integrating this technology into existing clinical practices. Successfully addressing these challenges could dramatically enhance the system's utility and reliability. Our vision extends beyond just semen analysis. The potential applications of this AI-based technology are vast and varied, ranging from disease diagnosis and pathology to clinical research, personalized health solutions, and even telemedicine. This expansion could revolutionize diagnostic and treatment processes across a wide array of medical fields. We are also cognizant of our study's limitations, including its external validity and technological dependence. These insights are invaluable, as they guide our future research efforts to further refine and expand the applications of AI in medical diagnostics and treatment. This research not only opens new avenues in reproductive health but also holds the promise of advancing medical science as a whole. As part of our future plans, we aim to implement this system in the near future, with plans to publish our findings. We also intend to merge artificial intelligence with biological insights to minimize errors and optimize semen analysis results. This dual approach will allow us to offer a semi-automatic device to our potential clients, ensuring accurate results and efficient patient file management for better follow-up and management. The completion of our startup is anticipated to open up new budget positions for our clients, facilitating the absorption and recruitment of graduates specializing in reproductive and developmental biology. This integration of cutting-edge technology and specialized human expertise is set to herald a new era in reproductive health diagnostics and treatment.