Claims:We claim,
1. A system for automated detection of malaria parasites based on the following elements:
A polymer substrate based ball lens microscope with a single broad spectrum light source enclosed within a non-conducting case. The whole unit being mounted on a cellphone using magnetic mounts, allowing photos to be captured of sub-micron specimens. A software application running on the cellphone captures images of the specimen and provides an interface for testers to input demographic and geological information. The photos thus captured are uploaded on a remote server along with additional information where it is analyzed for the presence of malaria parasites. The relevant results are relayed back to the concerned user and also logged onto the system for future analysis by researchers and health authorities.
2. A microscope as claimed in claim 1, based on said polymerised substance like polymer coated paper, flex papers, polyamide or thin metal used to support the said unit structure, providing it strength while remaining light weight.
3. A microscope as claimed in claim 1, containing the said single ball lens having a diameter in the range of 200 microns to 500 microns providing a magnification in the range of 500x to 1500x and a resolution of about 0.5 microns to 2 microns, enclosed in a changeable clip holder allowing usage of different magnification lens.
4. A microscope as claimed in claim 1, containing a said light module consisting of an LED-based light source working the visible light spectrum powered by a DC power source. A said condenser lens is used to focus the light source and prevent spherical aberrations.
5. A microscope as claimed in claim 1, containing said magnetic strips on the posterior side of the said microscope module, allowing it to be mounted on the said cellphones. The alignment with the said camera being performed by visual inspection.
6. A software application as claimed in claim 1, which allows the end-users to capture images of said concerned specimen, using the said computing device’s camera module, from the said microscope as claimed in claims 1 through 5, and also allowing the end-users to input demographic information including but not limited to age, sex, income band, food habits and geographic information including but not limited to GPS coordinates related to the said specimen. This information along with the captured image as claimed is then uploaded to the said remote server.
7. A remote system as claimed in claim 1, which receives images of said specimen from said remote clients, passes the input through a said band pass filter to remove Gaussian white noise, followed by a said segmentation module which segments the said input based on the presence of the said Romanowsky marker or Geimsa stain marker or Field’s stain marker.
8. A remote system as claimed in claim 1, which in some embodiments uses a learning based algorithm to detect the presence of malaria parasites in the said input according to said WHO guidelines, while in some other embodiments uses a contour detection based module as described for detection of the presence of malaria parasites in the said input.
9. A remote system as claimed in claim 1, which relays back the result as obtained from the detection module as claimed in claim 8 back to the user and in some embodiments also uses the information as input for its learning module.
10. A remote system as claimed in claim 1, which logs the relevant information received as part of the said specimen from the said remote client and maps the results on a said topological map for real-time analysis, and which provides an interface for real-time viewing of said logged information for monitoring and mitigation purposes. , Description:Microscopes are ubiquitous devices used in different facets of everyday life. One of the major uses of microscopes is in the field of biomedicine and biochemistry where they are extensively used for visual examination of sub-microscopic structures and lifeforms. As such a multitude of different types of microscopes has been developed which find usage in high precision areas. Traditionally, for biomedical purposes, a compound light microscope has been used. These microscopes use a plurality of glass lens to magnify sub-micron structures in the presence of a broad-spectrum high-intensity light source. An inherent feature of this kind of setups is fragility and high cost. They require high maintenance and skilled technicians for proper operation. These characteristics render them immovable and unfit for operations in harsh environments. A direct consequence of this being severe impediments in setting up diagnostic laboratories in rural areas. In our invention, we propose to overcome these difficulties, so that healthcare becomes inexpensive and less technologically complex.

In general, our present invention consists of a ball lens microscope attached to a fiber based substrate, along with a light module, consisting of a single point light source and a condenser lens for protection against spherical aberrations. The whole setup is enclosed within a non-conductive enclosure for protection and durability. Two magnetic strips are attached to back of the enclosure which acts as the mounts. The structure is attached in alignment to a cell phone's camera, which is used to capture images of the specimen mounted under the lens of the microscope. The captured image is then sent to a server for an automated analysis of the specimen properties, and the results are sent back to the end-device accordingly.

A more comprehensive understanding of the present system can be obtained with the detailed description given in accompaniment of the provided diagrams.

Figure 1, shows the frontal view of the single lens microscope developed as part of the invention. A flexible yet durable material like polymer, thin metal, polymer coated paper, flex papers or polyamide is used as the substrate of the microscope. A spherical ball lens, as shown in figure 1 and figure 2 as part of the microscope, is attached to the substrate which acts as the main magnifier. The ball lens has a diameter in the range of 200 microns to 2000 microns depending on the range of magnification needed. This provides a magnification factor of around 500x to 1500x. The microscope resolution is in the range of 0.5 to 2 microns. Each lens having a different diameter provide a different level of magnification allowing specimens to be observed at different view levels. In some embodiments, the microscope features an interchangeable lens mechanism allowing multiple magnification ranges in a single microscope.

The microscope also features a light module having a LED working in the white spectrum field or a colored LED with appropriate filters powered by a DC power source, for example, a battery source, which acts as the source of illumination for bright field microscopy. The light module also has a mounted condenser lens to help focus the light field emitted from the LED source while preventing spherical aberrations.

The sample slides are inserted into a slot in between the light source and the lens for viewing. A holding clip holds the lens in position while two separate handles provide adjustment capabilities for the lens. The device is aligned with the cell-phone camera by simple manual observation using a set of magnetic strip mounts. The sets of magnetic strips are attached to the microscope on its posterior side as shown in Figure 2 and Figure 3 for mounting the microscope on a cellphone or any other camera enabled computing device as shown in Figure 4 and 5.

In some embodiments, the microscope is to be used with Romanowsky stained blood samples for Plasmodium protozoan detection. In other embodiments, it can be used for samples stained with Geimsa stains or Field’s stain. An Application software is provided with the microscope to aid in the detection process which is described next.

The application software provides an interface for the user to capture multiple images of the slide, a comparative analysis is performed and the one with the best visual clarity, measured by visual inspection, as described by flow diagram 6, is uploaded to a remote server. Along with the blood sample owner’s demographic and geographic details, including but not limited to name, age, sex, income band, the number of immediate family members and GPS coordinates. The data transmission is performed over the internet preferably through a mobile network.

The image thus received, from the bright light microscope, on the server, as shown in figure 9, is first passed through a band pass filter to remove Gaussian white noise common in uncontrolled environments. The sample is then iteratively sharpened using an unsharp masking algorithm until the required level of sharpness is acquired. It is then passed through a mean shifter. The 3-channel image is then flattened to a single channel and finally passed through a segmentation module as shown in the flowchart in figure 8.

The segmentation module, as depicted in the flowchart in figure 8, segments the infected blood cells from the rest of the substrate based on the Romanowsky satin or Geimsa stain or Field’s stain marker as maybe the case for the specific embodiment. The parasite detection algorithm takes the segmented image as input and counts the number of infected cells as per WHO guidelines, discarding any partially visible cells at the frame boundaries on the right and bottom but including those on the top and left.

In some embodiments, a learning based algorithm is used to detect the infected cells, trained on datasets labeled by human beings. In some embodiments, an image processing algorithm based on hierarchical contour detection techniques is used as a follow-up step to the image segmentation module as described earlier and depicted in the flowchart in figure 8.

The results obtained from the detection phase are logged in the system along with its related data and is then relayed back to the end-user along with the enhanced segmented image as shown in Flow diagram 7. A probability score is also relayed back to the user signifying the level of confidence in the results. The details are then required to be further reviewed by qualified healthcare workers.

All test results are stored in the system database as described above as shown in Figure 9. The system allows the data to be reviewed by humans and compared with actual results, and after reviewing the data to be used as input for the learning module. This makes the system a semi-online learning module which gets more accurate the more it is used.

The GPS data obtained alongside the specimen is used to populate a topological map with real-time malaria infection cases. It also provides an interface for quick viewing of information logged into the system for each of the cases viewed on the map.