Artificial intelligence in visual field testing may help save vision

At the last American Academy of Optometry meeting in San Antonio, I presented a lecture entitled, “Learning from the mistakes of others in actual malpractice cases resulting in blindness and in cases of death.”

Here I share a recent case not included in the lecture but highly relevant to this topic.

Jerome Sherman

A 15-year-old female presented to an optometrist for the first time with a new symptom of flashing lights, some nausea and headaches. Her general health history was unremarkable. Visual acuity was 20/20 in each eye without correction. The external exam, including pupils and confrontation visual fields, was normal. The posterior segment exam was also noted as normal, including the discs, macula and peripheral retina.

A visual field screening was performed because of the “flashes/headaches.” Reliability was noted to be “marginal/questionable.”

As instructed, the patient returned in 1 year, now age 16, and complained of some focusing issues at distance in addition to the flashing lights and the headaches. With -0.25 D spheres, visual acuity was correctable to 20/20 OU.

As previously, the external and posterior segment exam was unremarkable. The listed diagnosis was myopia and paresis of accommodation.

The patient returned in about a year, at age 17, with the chief complaint of “lost glasses.” The history of flashes and headaches is listed in the EMR, but it is unclear whether this is just “cut and paste.” Again, best corrected visual acuity was 20/20, and the findings matched the previous exams.

The last visit to this optometrist occurred about a year later, at age 18, and all findings were the same as previously.

Observations

Note that the original field screening performed on the first visit (when the patient was age 15) for the flashes/headaches was described to be marginal/questionable but was never repeated on the three follow-up exams when the patient was 16, 17 and 18 years old. Confrontation visual fields were noted to be performed on all of the four visits and were recorded as normal.

Visual field screening on first visit. The yellow circles were drawn years after the exam.
Source: Jerome Sherman, OD, FAAO

A later review of the patient’s OB/GYN records noted “menstrual migraines.” A similar review of the pediatrician’s records noted a possible hearing deficit in the right ear, but a referral to ENT did not result in any imaging. None of the clinicians appear to have appropriately addressed the headaches.

Case continued

At age 19 years, 9 months following the optometrist’s last exam, the patient presented with family members to the ER and was diagnosed with “altered mental state” and “severe drowsiness, lethargy and acute blindness.”

CT and then an MRI revealed a large right parieto-occipital falcine meningioma that measured 8 cm x 7 cm. Craniotomy and resection of the tumor was performed in a 9-hour operation. Surgery removed 95% of the mass. A shunt followed, and gamma knife radiation was performed for the small remaining lesion.

Some minimal vision returned, and the patient and family noted that she was less lethargic and that “cognition was improving” somewhat, but this young lady will never be the same. The most recent visual acuity was reported as 20/50 in the right eye with a very limited field and counting fingers at 2 feet in the left eye. She uses a cane and is considering a seeing eye dog. Not surprisingly, she was also diagnosed as having posttraumatic stress disorder and severe anxiety.

This is a tragic case, and there is no denying that an earlier, timelier diagnosis could have been made by several of the patient’s doctors. Our specific interest relates to the ophthalmic care rendered.

What can we learn?

What can we learn from the mistakes made in this case?

First, it is important to note that although flashing lights are most often related to vitreal-retinal traction or prodromal migraine, pressure (from a mass) on any part of the visual pathway can result in flashing lights.

This sagittal MRI images both the right eye and the parieto-occipital mass. Note the size of the mass relative to the size of the globe.

Second, if you perform a visual field for the right reason (in this case because of the flashes/headaches) and the results are “marginal/questionable,” repeat the visual fields. This can be done on the same visit or on a repeat visit in a month or so. Also, consider threshold fields that take longer but typically yield far more information. In this case, the optometrist did the fields on the first visit but never repeated the fields on any of the subsequent three visits.

Third, do not rely on confrontation visual fields. Normal confrontations give the doctor and the patient a false sense of security. (Note that confrontation visual fields missed another brain tumor in a 15-year-old whose vision dropped from 20/40 and 20/20 to NLP in both eyes 1 year later. This case went to trial and resulted in a $9.2 million award.)

There is evidence in the screening visual field hinting at a problem in the brain — post-chiasmal on the right side in the parietal lobe. More specifically, there were three missed points in the inferior nasal quadrant in the right eye and four missed points in the inferior temporal quadrant in the left eye. These two quadrants correspond and suggest a brain lesion of some type beyond the chiasm on the right side of the brain.

Can technology help?

We hear about artificial intelligence (AI) nearly daily and recognize how AI will change our lives in the near future. This case and its tragic outcome cries out for AI to be incorporated into medicine but more specifically for us into eye care. A myriad of books, references and publications incorporate images and even models documenting all of the possible abnormal visual field patterns. Our knowledge of various visual field patterns and the corresponding locations within the retina, optic nerve and visual pathway of related abnormalities can be paired with AI.

With AI, the visual field performed nearly 4 years prior to the diagnosis of the meningioma would have been detected as a probable lesion post-chiasmal in the right parietal lobe.

Timely diagnosis and appropriate intervention would have prevented this tragic outcome. Is this not what we have devoted our professional lives to do?

In a future article I will provide tips on covering your bases to avoid a malpractice case such as this.

Disclosure: Sherman reports no relevant financial disclosures.

At the last American Academy of Optometry meeting in San Antonio, I presented a lecture entitled, “Learning from the mistakes of others in actual malpractice cases resulting in blindness and in cases of death.”

Here I share a recent case not included in the lecture but highly relevant to this topic.

Jerome Sherman

A 15-year-old female presented to an optometrist for the first time with a new symptom of flashing lights, some nausea and headaches. Her general health history was unremarkable. Visual acuity was 20/20 in each eye without correction. The external exam, including pupils and confrontation visual fields, was normal. The posterior segment exam was also noted as normal, including the discs, macula and peripheral retina.

A visual field screening was performed because of the “flashes/headaches.” Reliability was noted to be “marginal/questionable.”

As instructed, the patient returned in 1 year, now age 16, and complained of some focusing issues at distance in addition to the flashing lights and the headaches. With -0.25 D spheres, visual acuity was correctable to 20/20 OU.

As previously, the external and posterior segment exam was unremarkable. The listed diagnosis was myopia and paresis of accommodation.

The patient returned in about a year, at age 17, with the chief complaint of “lost glasses.” The history of flashes and headaches is listed in the EMR, but it is unclear whether this is just “cut and paste.” Again, best corrected visual acuity was 20/20, and the findings matched the previous exams.

The last visit to this optometrist occurred about a year later, at age 18, and all findings were the same as previously.

Observations

Note that the original field screening performed on the first visit (when the patient was age 15) for the flashes/headaches was described to be marginal/questionable but was never repeated on the three follow-up exams when the patient was 16, 17 and 18 years old. Confrontation visual fields were noted to be performed on all of the four visits and were recorded as normal.

Visual field screening on first visit. The yellow circles were drawn years after the exam.
Source: Jerome Sherman, OD, FAAO

A later review of the patient’s OB/GYN records noted “menstrual migraines.” A similar review of the pediatrician’s records noted a possible hearing deficit in the right ear, but a referral to ENT did not result in any imaging. None of the clinicians appear to have appropriately addressed the headaches.

Case continued

At age 19 years, 9 months following the optometrist’s last exam, the patient presented with family members to the ER and was diagnosed with “altered mental state” and “severe drowsiness, lethargy and acute blindness.”

CT and then an MRI revealed a large right parieto-occipital falcine meningioma that measured 8 cm x 7 cm. Craniotomy and resection of the tumor was performed in a 9-hour operation. Surgery removed 95% of the mass. A shunt followed, and gamma knife radiation was performed for the small remaining lesion.

Some minimal vision returned, and the patient and family noted that she was less lethargic and that “cognition was improving” somewhat, but this young lady will never be the same. The most recent visual acuity was reported as 20/50 in the right eye with a very limited field and counting fingers at 2 feet in the left eye. She uses a cane and is considering a seeing eye dog. Not surprisingly, she was also diagnosed as having posttraumatic stress disorder and severe anxiety.

This is a tragic case, and there is no denying that an earlier, timelier diagnosis could have been made by several of the patient’s doctors. Our specific interest relates to the ophthalmic care rendered.

PAGE BREAK

What can we learn?

What can we learn from the mistakes made in this case?

First, it is important to note that although flashing lights are most often related to vitreal-retinal traction or prodromal migraine, pressure (from a mass) on any part of the visual pathway can result in flashing lights.

This sagittal MRI images both the right eye and the parieto-occipital mass. Note the size of the mass relative to the size of the globe.

Second, if you perform a visual field for the right reason (in this case because of the flashes/headaches) and the results are “marginal/questionable,” repeat the visual fields. This can be done on the same visit or on a repeat visit in a month or so. Also, consider threshold fields that take longer but typically yield far more information. In this case, the optometrist did the fields on the first visit but never repeated the fields on any of the subsequent three visits.

Third, do not rely on confrontation visual fields. Normal confrontations give the doctor and the patient a false sense of security. (Note that confrontation visual fields missed another brain tumor in a 15-year-old whose vision dropped from 20/40 and 20/20 to NLP in both eyes 1 year later. This case went to trial and resulted in a $9.2 million award.)

There is evidence in the screening visual field hinting at a problem in the brain — post-chiasmal on the right side in the parietal lobe. More specifically, there were three missed points in the inferior nasal quadrant in the right eye and four missed points in the inferior temporal quadrant in the left eye. These two quadrants correspond and suggest a brain lesion of some type beyond the chiasm on the right side of the brain.

Can technology help?

We hear about artificial intelligence (AI) nearly daily and recognize how AI will change our lives in the near future. This case and its tragic outcome cries out for AI to be incorporated into medicine but more specifically for us into eye care. A myriad of books, references and publications incorporate images and even models documenting all of the possible abnormal visual field patterns. Our knowledge of various visual field patterns and the corresponding locations within the retina, optic nerve and visual pathway of related abnormalities can be paired with AI.

With AI, the visual field performed nearly 4 years prior to the diagnosis of the meningioma would have been detected as a probable lesion post-chiasmal in the right parietal lobe.

Timely diagnosis and appropriate intervention would have prevented this tragic outcome. Is this not what we have devoted our professional lives to do?

In a future article I will provide tips on covering your bases to avoid a malpractice case such as this.

Disclosure: Sherman reports no relevant financial disclosures.