Skip to main content

Table 3 Proposed Target Product Profile (TPP) for an IR-ML based system for entomological surveillance of malaria

From: Key considerations, target product profiles, and research gaps in the application of infrared spectroscopy and artificial intelligence for malaria surveillance and diagnosis

Characteristics

Entomological surveillance

Essential characteristics

Desired characteristics

Scope

Intended use

Mosquito identification; mosquito age-grading, detection of Plasmodium-infected mosquitoes and blood meal identity

One-stop platform for most desired entomological indicators; species IDs, age grading, blood meal identifications, infection detection and assessment of insecticide resistance status i.e. ability to predict resistance phenotype as per standard bioassays

Implementation level

Can be used in research laboratories and training centres

Used for research and routine surveillance services in laboratory and field settings

Types of spectrometers and durability

Bench-top units that require minimum electricity, and can operate 5 years or more with minimal maintenance

Off –the-shelf portable units with long-life battery; can be solar-powered; can operate for up to 10 years with minimal maintenance

Technical performance

Sensitivity relative to current methods [17, 18]

Plasmodium sporozoites: Sensitivity of > 90% with reference to ELISA

Plasmodium sporozoites: Sensitivity of > 90% with reference to PCR

Specificity relative to current methods [17, 18, 83]

Plasmodium sporozoites: Specificity of > 90% with reference to ELISA, Microscopy or PCR

Plasmodium sporozoites: Specificity of > 90% with reference to PCR

Resolution of predictions

Mosquitoes age: can classify young (e.g. 1–6 days) vs old (e.g. > 10 days) mosquitoes, with > 90% accuracy

Species identification: Can distinguish between members of a species complex (e.g. An. gambiae vs. An. arabiensis) with > 90% accuracy

Mosquito blood meals: Can distinguish between human blood meal from any other vertebrate blood meals, with > 90% accuracy

Mosquitoes age: Can classify mosquito ages chronologically at 2-day resolution (e.g. 1, 3, 5 days etc.) with > 90% accuracy

Species: Can identify all major malaria vectors (An. gambiae, An. arabiensis, An. coluzzii and An. funestus) with > 90% accuracy

Blood meal: Can identify most vertebrate blood meals (i.e. human, cattle, goat, chicken, dog etc.) even if the meals are mixed, and with > 90% accuracy

Lower limit of parasite detection

Plasmodium sporozoites: has performance equivalent to ELISA or PCR

Plasmodium sporozoites: has performance equivalent to PCR

Temperature and humidity stability

Functions and can be stored in ambient temperature; withstands frequent temperature rises to 35 °C for a long periods under diverse humidity

Functions and can be stored in ambient temperature; withstands frequent temperature rises to 45 °C for a long periods under diverse humidity

Operational aspects

Equipment & maintenance costs

Less than $ 30,000 per spectrometer; lasts > 5 years; costs < $100/yr to maintain

Less than $ 2000 per portable spectrometer; lasts > 5 yrs; Costs < $50/Yr to maintain

Sample handling costs

Costs < $0.1 per test

Costs < $0.01per test

Test duration

 < 5 min

 < 1 min

Sample preservation and storage

Requires freezing, desiccants, RNAlater

Uses must be less than 30 days old

Can use any preservation method

Samples can be older than 6 months as long as preserved fresh

Operational aspects

Human resources: Skills & training

Requires minimal training on sample handling, scanning & data/results interpretation

No more than 30 min training needed to use, conduct the tests and read results

Type of sample

Can analyse dried mosquito body parts

Can analyse either fresh or dried mosquito body parts

Reagents

No reagents needed except for cleaning the instruments or sample collection

No reagents needed except for cleaning the instruments or sample collection