BaLM / BaLM₁₀ by Prof. Michael Bach

Basic assessment of Light, Location & Motion perception

Assessing visual perception for ultra low acuity (below hand motion) is not covered by the usual acuity tests. Thus I developed this test for Retinal Implant. They kindly made it free for general use in 2019. The test and clinical results are described here:
Bach M, Wilke M, Wilhelm B, Zrenner E, Wilke R (2010) Basic quantitative assessment of visual performance in patients with very low vision. Invest Ophthalmol Vis Sci 51:1255–1260 [→PDF]. It has since been utilized in a number of studies.

New: BaLM incorporated in FrACT₁₀ → BaLM₁₀

The Light, Location and Motion tests are now available from within FrACT₁₀.

The test paradigm remains unchanged from that described in Bach M, Wilke M, Wilhelm B, Zrenner E, Wilke R (2010), though the Time test was left out due to its specificity for retinal implant devices and its lack of additional utility. Parameters can be adjusted in Settings▸BaLM.

If you want to try it out in FrACT₁₀, best first select TestingBaLM from Settings▸PRESETS for useful starting values, then run a BaLM test (using cursor keys to respond).

BaLM₁₀ Manual

Basic approach

Three basic dimensions of vision, namely light perception, light localization and motion perception are tested in a very straightforward way to assess ultra-low vision (worse than finger counting or hand movement). Stimuli are presented on a screen, announced by a sound, and the observer responds by pressing on of the cursor keys: → ↑ ← ↓.

Light perception is tested with a full-screen light flash, reaction is seen (↑) / not seen (↓)
Localisation is tested by starting with a huge fixation target and then adding a large wedge to it in one of four directions; response indicates one of the four: → ↑ ← ↓
Motion is tested with, yes, motion, of large disks on the screen in one of four directions; response indicates one of the four: → ↑ ← ↓

Settings▸PRESETS▸BaLM

General
- ISI. Inter-stimulus interval, in milliseconds. For the Light test it defines the dark phase before the flash; flash occurrence with 50% probability. Default: 1500 ms.
- Timeout. Maximal allowed response time, in seconds. After this time the trial is counted as a incorrect response. Default: 20 s.
- Distance. Observer distance, in centimeters. Default: 399 cm (which requires changing it), typical value would be 60 cm.
Light
- Light On. Duration of the flash, in milliseconds. Default: 200 ms.
- # trials. Number of trials for the 2-alternative condition (flash / no flash). Default: 32 (for BaLM, 24 suffice).
Location
- Diameter. Diameter of the circular fixation target seen before the wedge to be localized appears; in degrees visual angle. Default: 5°.
- Eccentricity. Extent of wedge; in degrees visual angle. Default: 15°.
- # trials. Number of trials for the 4-alternative localisation task. Default: 24.
Motion
- Diameter. Diameter of the moving disks in degrees visual angle. Default: 2°.
- Speed. Motion speed in degrees per second. Default: 3.3°/s.
- # trials. Number of trials for the 4-alternative motion task. This field is identical to the one of Localisation. Default: 24.

Result

The hit rate is the outcome variable of every test. It is visible on screen and also available for further processing with all of FrACT’s export capabilities.

Download options

For BaLM₁₀ go to FrACT₁₀, no need to download.

The original BaLM, though decades old, works fine in MacOS Sonoma (2024) and Windows 10/11 (Intel or ARM):

Windows stand-alone application: BaLM3.2.exe.zip.
On current Windows versions you will see a dialog “protected … unknown app … publisher…” or words to that effect. Need to choose “more info”, allow execution. This message is normal, it occurs because I do not pay to be a developer. This message appears only when opening BaLM for the very first time.
MacOS stand-alone application: BaLM3.2.app.zip.
Since MacOS 10.9: Must right click, open and confirm.