Total platelet counts (TPC) allow the determination of thrombocytosis, a risk factor for stroke, heart attack, and blood clots. TPC is generally performed at the laboratory by flow cytometry with laser scattering or impedance detection. Due to inaccuracies in quantification, clinical pathologists rely on microscopy counts using the Neubauer chamber technique as a quality control measure. In many health conditions, regular follow-up of coagulation risk is crucial, therefore point-of-care (POC) diagnosis simplifies these procedures, taking platelet counts to the bedside.

Spectroscopy has a high potential for reagent-less POC miniaturized technologies. However, platelet detection in blood by standard spectroscopy analysis is challenging, due to their small size, low number when compared to red blood cells, and low spectral contrast comparing to hemoglobin.

In this exploratory research, we show that is possible to perform TPC by advanced spectroscopy analysis using a new processing methodology with self-learning artificial intelligence. Results show that TPC can be measured by visible-near infrared spectroscopy above the detection limit of 121×10⁹ cells/L (R²=0.9922), tested within the data range of 53×10⁹ to 860×10⁹ cells/L in dog blood. These results open the possibility for spectroscopy as a diagnostic technology for detecting high levels of platelet counts, directly in whole blood, towards the rapid POC diagnosis of thrombocytosis and stroke prevention.