#	Pagina
attuale pagina	/open-h2020/projects/221158/results.html

Report

Teaser, summary, work performed and final results

Periodic Reporting for period 1 - Segmentum Imaging (Discovery and expansion of new markets for mobile digital pathology & cell measurement tool)

Teaser

Summary

The Challenge:
Pathology measures quantifying cell numbers, densities and morphologies are routine across academic research and medical diagnosis. These measures are extremely time consuming. The options for making these measurements on digital samples are either very simple and slow or extremely expensive.

Segmentum Imaging (SI) software allows much faster measures with no loss of accuracy in a manner convenient for researchers and medical staff. SI can analyse any pathological tissue measures and is available for use with a regular tablet device meaning that analysis is independent of any expensive equipment. SI facilitates routine measures faster than currently possible with no drop-in accuracy. Trials in the Neuropathology department of a large London hospital proved SI was able to deliver 80-90% decrease in time taken for measures such as cell or biomarker-counting, with built in options for randomised or systematic sampling to counteract measurement bias by users.
The SI App is designed to dramatically increase the speed and flexibility of these measures, allowing medical and scientific experts to increase both their output of large studies and conduct smaller projects that would otherwise be left aside.

Feasibility Study Objective:
The overall project objective of this feasibility study is to expand the scope of the software for other potential clinical and medicinal markets, focusing in particular on andrology (male fertility and microbiology). The study will enable the preparation of a detailed business plan outlining the technical requirements, design specifications, market acceptance requirements and route to market for commercialisation of Segmentum for expansion for commercial and academic research institutions. To achieve this the following tasks will be undertaken:

Work performed

There were four areas of focus for the Phase 1 investigation into new markets.
1. Academic Microbiology
There was considerable interest amongst academic microbiologists for a system of this nature for essentially the same reason as SI has received good feedback for pathology and anatomy uses. The issues of speed, cost, flexibility and ease of use were very well received by academic staff and students who currently have to make these measures.
2. Commercial Microbiology
The findings in commercial areas were more complex. Large-scale biotechnological showed little interest in our solution, due to the fact that they are the market targets of the large, high cost systems already available and therefore didnâ€™t see the need for a system such as SI. In contrast, smaller groups such as start-ups and CROâ€™s showed some interest due to its flexible, low-cost and quantitative nature. Also there was some interest in companies utilising bioreactors, as there is a significant issue with qualitative review of samples prior to large-scale growth being unsatisfactory, but no current quantitative fast system was available.
3. Medical Microbiology
Medical staff were almost universal in reporting that microscopy made up a tiny fraction of diagnostic methods in the modern biomedical laboratory, and that new techniques were steadily removing even these small uses.
It seems that nuclear magnetic resonance is the key developing technology for identification of viral coat proteins, and therefore critical in diagnosis of viral infections. Whereas genetic screening, involving bacterial plasmid sequencing and genomic marker identification, is the area of favoured development to identify bacterial infections. The figure that was reported by several diagnostic microbiologists was that â€œless than 5% of our work is down a microscopeâ€, which is declining year-on-year.
4. Diagnostic Andrology
Customer discovery amongst andrologists has revealed clear gaps in diagnosis methods. The current inability of andrologists to make accurate, evidence-based assessments of sperm morphology means that patients suffering from infertility must go through a long series of treatments until one succeeds. This process is inefficient, costly and traumatic, leading to many dropping out of treatment. Whilst early stage customer discovery has revealed this need, we would develop the base technology of SI to optimise for human make fertility diagnosis to allow fast, quantified assessment of spermatozoa biology, as well as the appropriate pricing level and method. We have obtained formal support from the UK national andrology body and collaborations with several UK centres of excellence for this project.
Evidently the optimal beachhead market for microbiology is a similar one to pathology, that is UK-based academic hands-on users 18-35yo. From this we could expand towards the same user groups internationally. From this there may be vertical adoption by senior staff or institutions, but the reluctance of senior staff to commit in interviews makes this hard to judge. For here a more measured approach to larger-scale adoption by institutions and commercial users would be more profitable.

Final results

For academic researchers, open access options have been developed such as QuPath, ImageJ and CellProfiler. These are free but with limited functionality and slow to use.
Within drug discovery and biomedical research specialist digital pathology equipment includes Visiopharm, Media Cybernetics. However, these are high capital cost, cause bottlenecks for access and either can only analyse preset measures (restricting application) or are very slow.
The typical process within diagnostic pathology laboratories is traditional and time consuming with manual counts performed using a microscope.

Our approach modernises the viewing of cells under a microscope during cell measures, vastly decreasing the time taken to undertake pathological studies with improved accuracy over manual counting. The system facilitates the overall accessibility of the data via direct download or cloud services as the user wishes. For use on any iOS device not simply single use hardware (PCs & laptops) often the cause of bottlenecks prevents the use of equipment bottlenecks. The app automatically tracks data generation to reduce upload delays and negate addition errors. The app does not require mobile access to function, but can make use of it, allowing data can be gathered and uploaded in transit and most importantly not limited to location and negates the need to upload data onto a centralised system database.
Currently the digital market provides a balance of cheap-and-minimal or optimised-but-expensive products. Which apply to the extreme ends of the world market. SI is developed to be affordable for any lab, especially those in emerging economies such as India, South America and Eastern Europe, and yet provide a toolset for counting, morphology and statistical measures both convenient to the user and of notably better quality than those offered by free services now.

Automation: Allows process from: â€œsample captured > measure > read > upload > data recordâ€ in a matter of minutes as opposed to days. Automated image segmentation ensures conscious and unconscious biases are avoided to increase accuracy. Similarly, with data collection; ensures data can be uploaded remotely and allows for the independent examination of results from any source worldwide.